You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@systemml.apache.org by de...@apache.org on 2016/05/05 01:13:54 UTC
[1/2] incubator-systemml git commit: [SYSTEMML-647] Replace
castAsScalar calls
Repository: incubator-systemml
Updated Branches:
refs/heads/master 7013910e0 -> 2da814574
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/wfundInputGenerator2.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/wfundInputGenerator2.dml b/src/test/scripts/applications/impute/wfundInputGenerator2.dml
index e6d302d..18d0ee2 100644
--- a/src/test/scripts/applications/impute/wfundInputGenerator2.dml
+++ b/src/test/scripts/applications/impute/wfundInputGenerator2.dml
@@ -124,7 +124,7 @@ disabled_known_values = disabled_known_values_extended;
is_free = matrix (1.0, rows = num_attrs, cols = 1);
for (i in 1:num_attrs) {
- j = castAsScalar (subtotals_tree [i, 1]);
+ j = as.scalar (subtotals_tree [i, 1]);
if (j > 0.0) {
is_free [j, 1] = 0.0 + zero;
} else {
@@ -137,12 +137,12 @@ num_frees = num_state_terms * num_frees_per_term;
CReps_block = matrix (0.0, rows = num_attrs, cols = num_frees_per_term);
index_free = 0;
for (i in 1:num_attrs) {
- if (castAsScalar (is_free [i, 1]) == 1.0) {
+ if (as.scalar (is_free [i, 1]) == 1.0) {
index_free = index_free + 1;
j = i;
while (j > 0.0) {
CReps_block [j, index_free] = 1.0 + zero;
- j = castAsScalar (subtotals_tree [j, 1]);
+ j = as.scalar (subtotals_tree [j, 1]);
} } }
CReps = matrix (0.0, rows = (num_terms * num_attrs), cols = num_frees);
@@ -221,7 +221,7 @@ RegresFactorDefault = matrix (0.0, rows = (num_reg_eqs * num_factors), cols = 1)
for (t in 2 : num_state_terms) {
for (i in 1 : num_attrs) {
reg_index = ((t-1) * num_attrs - 1 + i) * num_factors;
- agg = castAsScalar (subtotals_tree [i, 1]);
+ agg = as.scalar (subtotals_tree [i, 1]);
if (i <= 18 & agg > 0)
{
RegresValueMap [reg_index + 1, (t-1) * num_attrs + i ] = -1.0 + zero; # 1st factor: -x[t]
@@ -256,8 +256,8 @@ for (t in 2 : num_state_terms) {
for (t1 in (num_state_terms + 1) : num_terms) {
t2 = t1 - num_state_terms;
for (i in 1 : num_attrs) {
- if ((i <= num_observed_attrs & t2 <= num_known_terms & castAsScalar (disabled_known_values [i, t2]) == 0.0) |
- (i > num_observed_attrs & castAsScalar (subtotals_tree [i, 1]) > 0.0))
+ if ((i <= num_observed_attrs & t2 <= num_known_terms & as.scalar (disabled_known_values [i, t2]) == 0.0) |
+ (i > num_observed_attrs & as.scalar (subtotals_tree [i, 1]) > 0.0))
{
reg_index = ((t1 - 1) * num_attrs - 1 + i) * num_factors;
RegresValueMap [reg_index + 1, (t1 - 1) * num_attrs + i] = -1.0 + zero; # 1st factor: -y[t]
@@ -291,7 +291,7 @@ RegresCoeffDefault = matrix (0.0, rows = (num_reg_eqs * num_factors), cols = 1);
for (t in 2 : num_state_terms) {
for (i in 1 : num_observed_attrs) {
- if (castAsScalar (subtotals_tree [i, 1]) > 0.0) {
+ if (as.scalar (subtotals_tree [i, 1]) > 0.0) {
param_1 = 3 * i - 1;
param_2 = 3 * i;
param_3 = 3 * i + 1;
@@ -308,7 +308,7 @@ for (t in 2 : num_state_terms) {
RegresCoeffDefault [reg_index + 4, 1] = 1.0 + zero;
for (i in (num_observed_attrs + 1) : num_attrs) {
- if (castAsScalar (subtotals_tree [i, 1]) > 0.0) {
+ if (as.scalar (subtotals_tree [i, 1]) > 0.0) {
reg_index = ((t-1) * num_attrs - 1 + i) * num_factors;
RegresCoeffDefault [reg_index + 1, 1] = 1.0 + zero;
RegresCoeffDefault [reg_index + 2, 1] = 1.0 + zero;
@@ -331,8 +331,8 @@ for (t in 2 : num_state_terms) {
for (t1 in (num_state_terms + 1) : num_terms) {
t2 = t1 - num_state_terms;
for (i in 1 : num_attrs) {
- if ((i <= num_observed_attrs & t2 <= num_known_terms & castAsScalar (disabled_known_values [i, t2]) == 0.0) |
- (i > num_observed_attrs & castAsScalar (subtotals_tree [i, 1]) > 0.0))
+ if ((i <= num_observed_attrs & t2 <= num_known_terms & as.scalar (disabled_known_values [i, t2]) == 0.0) |
+ (i > num_observed_attrs & as.scalar (subtotals_tree [i, 1]) > 0.0))
{
reg_index = ((t1 - 1) * num_attrs - 1 + i) * num_factors;
RegresCoeffDefault [reg_index + 1, 1] = 1.0 + zero;
@@ -352,7 +352,7 @@ RegresParamMap [reg_index + 1, param] = 1.0 + zero;
RegresCoeffDefault [reg_index + 2, 1 ] = 0.0 + zero;
for (i in 1 : num_observed_attrs) {
- agg = castAsScalar (subtotals_tree [i, 1]);
+ agg = as.scalar (subtotals_tree [i, 1]);
if (agg >= 0.0)
{
param = 3 * i - 1;
@@ -397,7 +397,7 @@ for (i in 1 : num_attrs)
scale_factor = 1.0;
if (i <= num_observed_attrs) {
### CORRECTION FOR OBSERVED ATTRIBUTES:
- attribute_size_i = castAsScalar (attribute_size [i, 1]);
+ attribute_size_i = as.scalar (attribute_size [i, 1]);
scale_factor = sqrt (attribute_size_i / max_attr_size) * 0.999 + 0.001;
}
for (t in 1 : num_terms) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/linearLogReg/LinearLogReg.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/linearLogReg/LinearLogReg.dml b/src/test/scripts/applications/linearLogReg/LinearLogReg.dml
index cf2f7ad..4c84158 100644
--- a/src/test/scripts/applications/linearLogReg/LinearLogReg.dml
+++ b/src/test/scripts/applications/linearLogReg/LinearLogReg.dml
@@ -90,7 +90,7 @@ while(!converge) {
norm_grad = sqrt(sum(grad*grad))
print("-- Outer Iteration = " + iter)
- objScalar = castAsScalar(obj)
+ objScalar = as.scalar(obj)
print(" Iterations = " + iter + ", Objective = " + objScalar + ", Gradient Norm = " + norm_grad)
# SOLVE TRUST REGION SUB-PROBLEM
@@ -108,23 +108,23 @@ while(!converge) {
alpha_deno = t(d) %*% Hd
alpha = norm_r2 / alpha_deno
- s = s + castAsScalar(alpha) * d
- os = os + castAsScalar(alpha) * od
+ s = s + as.scalar(alpha) * d
+ os = os + as.scalar(alpha) * od
sts = t(s) %*% s
delta2 = delta*delta
- stsScalar = castAsScalar(sts)
+ stsScalar = as.scalar(sts)
shouldBreak = FALSE; # to mimic "break" in the following 'if' condition
if (stsScalar > delta2) {
print(" --- cg reaches trust region boundary")
- s = s - castAsScalar(alpha) * d
- os = os - castAsScalar(alpha) * od
+ s = s - as.scalar(alpha) * d
+ os = os - as.scalar(alpha) * od
std = t(s) %*% d
dtd = t(d) %*% d
sts = t(s) %*% s
rad = sqrt(std*std + dtd*(delta2 - sts))
- stdScalar = castAsScalar(std)
+ stdScalar = as.scalar(std)
if(stdScalar >= 0) {
tau = (delta2 - sts)/(std + rad)
}
@@ -132,9 +132,9 @@ while(!converge) {
tau = (rad - std)/dtd
}
- s = s + castAsScalar(tau) * d
- os = os + castAsScalar(tau) * od
- r = r - castAsScalar(tau) * Hd
+ s = s + as.scalar(tau) * d
+ os = os + as.scalar(tau) * od
+ r = r - as.scalar(tau) * Hd
#break
shouldBreak = TRUE;
@@ -143,7 +143,7 @@ while(!converge) {
}
if (!shouldBreak) {
- r = r - castAsScalar(alpha) * Hd
+ r = r - as.scalar(alpha) * Hd
old_norm_r2 = norm_r2
norm_r2 = sum(r*r)
beta = norm_r2/old_norm_r2
@@ -164,10 +164,10 @@ while(!converge) {
objnew = 0.5 * t(wnew) %*% wnew + C * sum(-log(logisticnew))
actred = (obj - objnew)
- actredScalar = castAsScalar(actred)
+ actredScalar = as.scalar(actred)
rho = actred / qk
- qkScalar = castAsScalar(qk)
- rhoScalar = castAsScalar(rho);
+ qkScalar = as.scalar(qk)
+ rhoScalar = as.scalar(rho);
snorm = sqrt(sum( s * s ))
print(" Actual = " + actredScalar)
print(" Predicted = " + qkScalar)
@@ -176,12 +176,12 @@ while(!converge) {
delta = min(delta, snorm)
}
alpha2 = objnew - obj - gs
- alpha2Scalar = castAsScalar(alpha2)
+ alpha2Scalar = as.scalar(alpha2)
if (alpha2Scalar <= 0) {
alpha = sigma3*e
}
else {
- ascalar = max(sigma1, -0.5*castAsScalar(gs)/alpha2Scalar)
+ ascalar = max(sigma1, -0.5*as.scalar(gs)/alpha2Scalar)
alpha = ascalar*e
}
@@ -195,7 +195,7 @@ while(!converge) {
obj = objnew
}
- alphaScalar = castAsScalar(alpha)
+ alphaScalar = as.scalar(alpha)
if (rhoScalar < eta0){
delta = min(max( alphaScalar , sigma1) * snorm, sigma2 * delta )
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/linearLogReg/LinearLogReg.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/linearLogReg/LinearLogReg.pydml b/src/test/scripts/applications/linearLogReg/LinearLogReg.pydml
index 1a9e769..188cd42 100644
--- a/src/test/scripts/applications/linearLogReg/LinearLogReg.pydml
+++ b/src/test/scripts/applications/linearLogReg/LinearLogReg.pydml
@@ -91,7 +91,7 @@ while(!converge):
norm_grad = sqrt(sum(grad*grad))
print("-- Outer Iteration = " + iter)
- objScalar = castAsScalar(obj)
+ objScalar = scalar(obj)
print(" Iterations = " + iter + ", Objective = " + objScalar + ", Gradient Norm = " + norm_grad)
# SOLVE TRUST REGION SUB-PROBLEM
@@ -109,37 +109,37 @@ while(!converge):
alpha_deno = dot(transpose(d), Hd)
alpha = norm_r2 / alpha_deno
- s = s + castAsScalar(alpha) * d
- os = os + castAsScalar(alpha) * od
+ s = s + scalar(alpha) * d
+ os = os + scalar(alpha) * od
sts = dot(transpose(s), s)
delta2 = delta*delta
- stsScalar = castAsScalar(sts)
+ stsScalar = scalar(sts)
shouldBreak = False # to mimic "break" in the following 'if' condition
if (stsScalar > delta2):
print(" --- cg reaches trust region boundary")
- s = s - castAsScalar(alpha) * d
- os = os - castAsScalar(alpha) * od
+ s = s - scalar(alpha) * d
+ os = os - scalar(alpha) * od
std = dot(transpose(s), d)
dtd = dot(transpose(d), d)
sts = dot(transpose(s), s)
rad = sqrt(std*std + dtd*(delta2 - sts))
- stdScalar = castAsScalar(std)
+ stdScalar = scalar(std)
if(stdScalar >= 0):
tau = (delta2 - sts)/(std + rad)
else:
tau = (rad - std)/dtd
- s = s + castAsScalar(tau) * d
- os = os + castAsScalar(tau) * od
- r = r - castAsScalar(tau) * Hd
+ s = s + scalar(tau) * d
+ os = os + scalar(tau) * od
+ r = r - scalar(tau) * Hd
#break
shouldBreak = True
innerconverge = True
if (!shouldBreak):
- r = r - castAsScalar(alpha) * Hd
+ r = r - scalar(alpha) * Hd
old_norm_r2 = norm_r2
norm_r2 = sum(r*r)
beta = norm_r2/old_norm_r2
@@ -159,10 +159,10 @@ while(!converge):
objnew = dot((0.5 * transpose(wnew)), wnew) + C * sum(-log(logisticnew))
actred = (obj - objnew)
- actredScalar = castAsScalar(actred)
+ actredScalar = scalar(actred)
rho = actred / qk
- qkScalar = castAsScalar(qk)
- rhoScalar = castAsScalar(rho)
+ qkScalar = scalar(qk)
+ rhoScalar = scalar(rho)
snorm = sqrt(sum( s * s ))
print(" Actual = " + actredScalar)
print(" Predicted = " + qkScalar)
@@ -170,11 +170,11 @@ while(!converge):
if (iter==0):
delta = min(delta, snorm)
alpha2 = objnew - obj - gs
- alpha2Scalar = castAsScalar(alpha2)
+ alpha2Scalar = scalar(alpha2)
if (alpha2Scalar <= 0):
alpha = sigma3*e
else:
- ascalar = max(sigma1, -0.5*castAsScalar(gs)/alpha2Scalar)
+ ascalar = max(sigma1, -0.5*scalar(gs)/alpha2Scalar)
alpha = ascalar*e
if (rhoScalar > eta0):
@@ -185,7 +185,7 @@ while(!converge):
logisticD = logisticnew * (1 - logisticnew)
obj = objnew
- alphaScalar = castAsScalar(alpha)
+ alphaScalar = scalar(alpha)
if (rhoScalar < eta0):
delta = min(max( alphaScalar , sigma1) * snorm, sigma2 * delta )
else:
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/linear_regression/LinearRegression.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/linear_regression/LinearRegression.dml b/src/test/scripts/applications/linear_regression/LinearRegression.dml
index e02c53d..5720b8b 100644
--- a/src/test/scripts/applications/linear_regression/LinearRegression.dml
+++ b/src/test/scripts/applications/linear_regression/LinearRegression.dml
@@ -40,7 +40,7 @@ max_iteration = 3;
i = 0;
while(i < max_iteration) {
q = ((t(V) %*% (V %*% p)) + eps * p);
- alpha = norm_r2 / castAsScalar(t(p) %*% q);
+ alpha = norm_r2 / as.scalar(t(p) %*% q);
w = w + alpha * p;
old_norm_r2 = norm_r2;
r = r + alpha * q;
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/m-svm/m-svm.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/m-svm/m-svm.dml b/src/test/scripts/applications/m-svm/m-svm.dml
index bbf5acc..7b2828f 100644
--- a/src/test/scripts/applications/m-svm/m-svm.dml
+++ b/src/test/scripts/applications/m-svm/m-svm.dml
@@ -136,7 +136,7 @@ if(check_X == 0){
debug_str = "# Class, Iter, Obj"
for(iter_class in 1:ncol(debug_mat)){
for(iter in 1:nrow(debug_mat)){
- obj = castAsScalar(debug_mat[iter, iter_class])
+ obj = as.scalar(debug_mat[iter, iter_class])
if(obj != -1)
debug_str = append(debug_str, iter_class + "," + iter + "," + obj)
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/m-svm/m-svm.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/m-svm/m-svm.pydml b/src/test/scripts/applications/m-svm/m-svm.pydml
index 348f599..fc3669c 100644
--- a/src/test/scripts/applications/m-svm/m-svm.pydml
+++ b/src/test/scripts/applications/m-svm/m-svm.pydml
@@ -129,7 +129,7 @@ else:
debug_str = "# Class, Iter, Obj"
for(iter_class in 1:ncol(debug_mat)):
for(iter in 1:nrow(debug_mat)):
- obj = castAsScalar(debug_mat[iter, iter_class])
+ obj = scalar(debug_mat[iter, iter_class])
if(obj != -1):
debug_str = append(debug_str, iter_class + "," + iter + "," + obj)
save(debug_str, $Log)
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/mdabivar/MDABivariateStats.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/mdabivar/MDABivariateStats.dml b/src/test/scripts/applications/mdabivar/MDABivariateStats.dml
index 56163ad..5bb980c 100644
--- a/src/test/scripts/applications/mdabivar/MDABivariateStats.dml
+++ b/src/test/scripts/applications/mdabivar/MDABivariateStats.dml
@@ -65,9 +65,9 @@ mn = colMins(D)
num_distinct_values = mx-mn+1
max_num_distinct_values = 0
for(i1 in 1:nrow(feature_indices)){
- feature_index1 = castAsScalar(feature_indices[i1,1])
- num = castAsScalar(num_distinct_values[1,feature_index1])
- if(castAsScalar(feature_measurement_levels[i1,1]) == 0 & num >= max_num_distinct_values){
+ feature_index1 = as.scalar(feature_indices[i1,1])
+ num = as.scalar(num_distinct_values[1,feature_index1])
+ if(as.scalar(feature_measurement_levels[i1,1]) == 0 & num >= max_num_distinct_values){
max_num_distinct_values = num
}
}
@@ -107,8 +107,8 @@ if(label_measurement_level == 0){
}
parfor(i3 in 1:nrow(feature_indices), check=0){
- feature_index2 = castAsScalar(feature_indices[i3,1])
- feature_measurement_level = castAsScalar(feature_measurement_levels[i3,1])
+ feature_index2 = as.scalar(feature_indices[i3,1])
+ feature_measurement_level = as.scalar(feature_measurement_levels[i3,1])
feature = D[,feature_index2]
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/mdabivar/MDABivariateStats.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/mdabivar/MDABivariateStats.pydml b/src/test/scripts/applications/mdabivar/MDABivariateStats.pydml
index 7fbc101..134eb4b 100644
--- a/src/test/scripts/applications/mdabivar/MDABivariateStats.pydml
+++ b/src/test/scripts/applications/mdabivar/MDABivariateStats.pydml
@@ -64,9 +64,9 @@ mn = colMins(D)
num_distinct_values = mx-mn+1
max_num_distinct_values = 0
for(i1 in 1:nrow(feature_indices)):
- feature_index1 = castAsScalar(feature_indices[i1,1])
- num = castAsScalar(num_distinct_values[1,feature_index1])
- if(castAsScalar(feature_measurement_levels[i1,1]) == 0 & num >= max_num_distinct_values):
+ feature_index1 = scalar(feature_indices[i1,1])
+ num = scalar(num_distinct_values[1,feature_index1])
+ if(scalar(feature_measurement_levels[i1,1]) == 0 & num >= max_num_distinct_values):
max_num_distinct_values = num
distinct_label_values = full(0, rows=1, cols=1)
contingencyTableSz = 1
@@ -99,8 +99,8 @@ if(label_measurement_level == 0):
featureValues[label_index,i2] = i2-labelCorrection
parfor(i3 in 1:nrow(feature_indices), check=0):
- feature_index2 = castAsScalar(feature_indices[i3,1])
- feature_measurement_level = castAsScalar(feature_measurement_levels[i3,1])
+ feature_index2 = scalar(feature_indices[i3,1])
+ feature_measurement_level = scalar(feature_measurement_levels[i3,1])
feature = D[,feature_index2]
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_bivariate0.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_bivariate0.dml b/src/test/scripts/applications/parfor/parfor_bivariate0.dml
index 341a0b1..662b745 100644
--- a/src/test/scripts/applications/parfor/parfor_bivariate0.dml
+++ b/src/test/scripts/applications/parfor/parfor_bivariate0.dml
@@ -67,16 +67,16 @@ cat_vars = matrix(0, rows=maxC, cols=numPairs);
for( i in 1:s1size ) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
#print("a1="+a1);
for( j in 1:s2size ) {
pairID = (i-1)*s2size+j;
#print("ID="+pairID+"(i="+i+",j="+j+")");
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
#print("a2="+a2);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_bivariate1.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_bivariate1.dml b/src/test/scripts/applications/parfor/parfor_bivariate1.dml
index aa40d7c..42a956d 100644
--- a/src/test/scripts/applications/parfor/parfor_bivariate1.dml
+++ b/src/test/scripts/applications/parfor/parfor_bivariate1.dml
@@ -67,14 +67,14 @@ cat_vars = matrix(0, rows=maxC, cols=numPairs);
parfor( i in 1:s1size, par=4, mode=LOCAL, check=0, opt=NONE) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, par=4, mode=LOCAL, check=0, opt=NONE) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
if (k1 == k2) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_bivariate2.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_bivariate2.dml b/src/test/scripts/applications/parfor/parfor_bivariate2.dml
index d0734a9..8ee59b7 100644
--- a/src/test/scripts/applications/parfor/parfor_bivariate2.dml
+++ b/src/test/scripts/applications/parfor/parfor_bivariate2.dml
@@ -67,14 +67,14 @@ cat_vars = matrix(0, rows=maxC, cols=numPairs);
parfor( i in 1:s1size, par=4, mode=LOCAL, check=0, opt=NONE) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, par=4, mode=REMOTE_MR, check=0, opt=NONE) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
if (k1 == k2) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_bivariate3.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_bivariate3.dml b/src/test/scripts/applications/parfor/parfor_bivariate3.dml
index 990a6fe..f5c43ef 100644
--- a/src/test/scripts/applications/parfor/parfor_bivariate3.dml
+++ b/src/test/scripts/applications/parfor/parfor_bivariate3.dml
@@ -67,14 +67,14 @@ cat_vars = matrix(0, rows=maxC, cols=numPairs);
parfor( i in 1:s1size, par=4, mode=REMOTE_MR, check=0, opt=NONE) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, par=4, mode=LOCAL, check=0, opt=NONE) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
if (k1 == k2) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_bivariate4.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_bivariate4.dml b/src/test/scripts/applications/parfor/parfor_bivariate4.dml
index a19957f..c2e78d5 100644
--- a/src/test/scripts/applications/parfor/parfor_bivariate4.dml
+++ b/src/test/scripts/applications/parfor/parfor_bivariate4.dml
@@ -70,14 +70,14 @@ cat_vars = matrix(0, rows=maxC, cols=numPairs);
parfor( i in 1:s1size, check=0) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, check=0) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
if (k1 == k2) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_univariate0.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_univariate0.dml b/src/test/scripts/applications/parfor/parfor_univariate0.dml
index 2a6a9c5..78397cb 100644
--- a/src/test/scripts/applications/parfor/parfor_univariate0.dml
+++ b/src/test/scripts/applications/parfor/parfor_univariate0.dml
@@ -77,7 +77,7 @@ else {
# project out the i^th column
F = A[,i];
- kind = castAsScalar(K[1,i]);
+ kind = as.scalar(K[1,i]);
if ( kind == 1 ) {
print("[" + i + "] Scale");
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_univariate1.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_univariate1.dml b/src/test/scripts/applications/parfor/parfor_univariate1.dml
index 1f120ef..64ee633 100644
--- a/src/test/scripts/applications/parfor/parfor_univariate1.dml
+++ b/src/test/scripts/applications/parfor/parfor_univariate1.dml
@@ -77,7 +77,7 @@ else {
# project out the i^th column
F = A[,i];
- kind = castAsScalar(K[1,i]);
+ kind = as.scalar(K[1,i]);
if ( kind == 1 ) {
print("[" + i + "] Scale");
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/parfor/parfor_univariate4.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/parfor/parfor_univariate4.dml b/src/test/scripts/applications/parfor/parfor_univariate4.dml
index 8953c64..465cb8f 100644
--- a/src/test/scripts/applications/parfor/parfor_univariate4.dml
+++ b/src/test/scripts/applications/parfor/parfor_univariate4.dml
@@ -77,7 +77,7 @@ else {
# project out the i^th column
F = A[,i];
- kind = castAsScalar(K[1,i]);
+ kind = as.scalar(K[1,i]);
if ( kind == 1 ) {
print("[" + i + "] Scale");
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/validation/CV_LogisticRegression.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/validation/CV_LogisticRegression.dml b/src/test/scripts/applications/validation/CV_LogisticRegression.dml
index e890b80..d91a3f7 100644
--- a/src/test/scripts/applications/validation/CV_LogisticRegression.dml
+++ b/src/test/scripts/applications/validation/CV_LogisticRegression.dml
@@ -165,7 +165,7 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
norm_grad = sqrt(sum(grad*grad))
#print("-- Outer Iteration = " + iter)
- objScalar = castAsScalar(obj)
+ objScalar = as.scalar(obj)
#print(" Iterations = " + iter + ", Objective = " + objScalar + ", Gradient Norm = " + norm_grad)
# SOLVE TRUST REGION SUB-PROBLEM
@@ -183,23 +183,23 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
alpha_deno = t(d) %*% Hd
alpha = norm_r2 / alpha_deno
- s = s + castAsScalar(alpha) * d
- os = os + castAsScalar(alpha) * od
+ s = s + as.scalar(alpha) * d
+ os = os + as.scalar(alpha) * od
sts = t(s) %*% s
delta2 = delta*delta
- stsScalar = castAsScalar(sts)
+ stsScalar = as.scalar(sts)
shouldBreak = FALSE; # to mimic "break" in the following 'if' condition
if (stsScalar > delta2) {
#print(" --- cg reaches trust region boundary")
- s = s - castAsScalar(alpha) * d
- os = os - castAsScalar(alpha) * od
+ s = s - as.scalar(alpha) * d
+ os = os - as.scalar(alpha) * od
std = t(s) %*% d
dtd = t(d) %*% d
sts = t(s) %*% s
rad = sqrt(std*std + dtd*(delta2 - sts))
- stdScalar = castAsScalar(std)
+ stdScalar = as.scalar(std)
tau = 0; #TODO
if(stdScalar >= 0) {
@@ -209,9 +209,9 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
tau = (rad - std)/dtd
}
- s = s + castAsScalar(tau) * d
- os = os + castAsScalar(tau) * od
- r = r - castAsScalar(tau) * Hd
+ s = s + as.scalar(tau) * d
+ os = os + as.scalar(tau) * od
+ r = r - as.scalar(tau) * Hd
#break
shouldBreak = TRUE;
@@ -220,7 +220,7 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
}
if (!shouldBreak) {
- r = r - castAsScalar(alpha) * Hd
+ r = r - as.scalar(alpha) * Hd
old_norm_r2 = norm_r2
norm_r2 = sum(r*r)
beta = norm_r2/old_norm_r2
@@ -241,10 +241,10 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
objnew = 0.5 * t(wnew) %*% wnew + C * sum(-log(logisticnew))
actred = (obj - objnew)
- actredScalar = castAsScalar(actred)
+ actredScalar = as.scalar(actred)
rho = actred / qk
- qkScalar = castAsScalar(qk)
- rhoScalar = castAsScalar(rho);
+ qkScalar = as.scalar(qk)
+ rhoScalar = as.scalar(rho);
snorm = sqrt(sum( s * s ))
#print(" Actual = " + actredScalar)
@@ -254,12 +254,12 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
delta = min(delta, snorm)
}
alpha2 = objnew - obj - gs
- alpha2Scalar = castAsScalar(alpha2)
+ alpha2Scalar = as.scalar(alpha2)
if (alpha2Scalar <= 0) {
alpha = sigma3*e
}
else {
- ascalar = max(sigma1, -0.5*castAsScalar(gs)/alpha2Scalar)
+ ascalar = max(sigma1, -0.5*as.scalar(gs)/alpha2Scalar)
alpha = ascalar*e
}
@@ -273,7 +273,7 @@ logisticRegression = function (Matrix[double] X, Matrix[double] y, Integer in_in
obj = objnew
}
- alphaScalar = castAsScalar(alpha)
+ alphaScalar = as.scalar(alpha)
if (rhoScalar < eta0){
delta = min(max( alphaScalar , sigma1) * snorm, sigma2 * delta )
}
@@ -342,7 +342,7 @@ scoreLogRegModel = function (Matrix[double] X_train, Matrix[double] y_train, Mat
b = 0.0;
if (nrow (w) > num_features) {
- b = castAsScalar (w [num_features + 1, 1]);
+ b = as.scalar (w [num_features + 1, 1]);
}
# TRAINING DATA - ESTIMATE PROBABILITIES:
@@ -472,56 +472,56 @@ printFoldStatistics = function (Matrix[double] stats ) return( Integer err )
stats_min = round (colMins (stats) * 10000.0) / 10000.0;
/*
print ("Training Data, Model-estimated statistics:");
- print (" True Positives: Min = " + castAsScalar (stats_min [1, 1]) + ", Avg = " + castAsScalar (stats_avg [1, 1]) + ", Max = " + castAsScalar (stats_max [1, 1]));
- print (" False Positives: Min = " + castAsScalar (stats_min [1, 2]) + ", Avg = " + castAsScalar (stats_avg [1, 2]) + ", Max = " + castAsScalar (stats_max [1, 2]));
- print (" True Negatives: Min = " + castAsScalar (stats_min [1, 3]) + ", Avg = " + castAsScalar (stats_avg [1, 3]) + ", Max = " + castAsScalar (stats_max [1, 3]));
- print (" False Negatives: Min = " + castAsScalar (stats_min [1, 4]) + ", Avg = " + castAsScalar (stats_avg [1, 4]) + ", Max = " + castAsScalar (stats_max [1, 4]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 5]) + ", Avg = " + castAsScalar (stats_avg [1, 5]) + ", Max = " + castAsScalar (stats_max [1, 5]));
- print ("Recall (Sensit-y)%: Min = " + castAsScalar (stats_min [1, 6]) + ", Avg = " + castAsScalar (stats_avg [1, 6]) + ", Max = " + castAsScalar (stats_max [1, 6]));
- print (" Specificity %: Min = " + castAsScalar (stats_min [1, 7]) + ", Avg = " + castAsScalar (stats_avg [1, 7]) + ", Max = " + castAsScalar (stats_max [1, 7]));
- print (" Value - Cost: Min = " + castAsScalar (stats_min [1, 8]) + ", Avg = " + castAsScalar (stats_avg [1, 8]) + ", Max = " + castAsScalar (stats_max [1, 8]));
+ print (" True Positives: Min = " + as.scalar (stats_min [1, 1]) + ", Avg = " + as.scalar (stats_avg [1, 1]) + ", Max = " + as.scalar (stats_max [1, 1]));
+ print (" False Positives: Min = " + as.scalar (stats_min [1, 2]) + ", Avg = " + as.scalar (stats_avg [1, 2]) + ", Max = " + as.scalar (stats_max [1, 2]));
+ print (" True Negatives: Min = " + as.scalar (stats_min [1, 3]) + ", Avg = " + as.scalar (stats_avg [1, 3]) + ", Max = " + as.scalar (stats_max [1, 3]));
+ print (" False Negatives: Min = " + as.scalar (stats_min [1, 4]) + ", Avg = " + as.scalar (stats_avg [1, 4]) + ", Max = " + as.scalar (stats_max [1, 4]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 5]) + ", Avg = " + as.scalar (stats_avg [1, 5]) + ", Max = " + as.scalar (stats_max [1, 5]));
+ print ("Recall (Sensit-y)%: Min = " + as.scalar (stats_min [1, 6]) + ", Avg = " + as.scalar (stats_avg [1, 6]) + ", Max = " + as.scalar (stats_max [1, 6]));
+ print (" Specificity %: Min = " + as.scalar (stats_min [1, 7]) + ", Avg = " + as.scalar (stats_avg [1, 7]) + ", Max = " + as.scalar (stats_max [1, 7]));
+ print (" Value - Cost: Min = " + as.scalar (stats_min [1, 8]) + ", Avg = " + as.scalar (stats_avg [1, 8]) + ", Max = " + as.scalar (stats_max [1, 8]));
print (" ");
if (1==1) {
print(" ")
}
*/
print ("Training Data, Label comparison statistics:");
- print (" True Positives: Min = " + castAsScalar (stats_min [1, 11]) + ", Avg = " + castAsScalar (stats_avg [1, 11]) + ", Max = " + castAsScalar (stats_max [1, 11]));
- print (" False Positives: Min = " + castAsScalar (stats_min [1, 12]) + ", Avg = " + castAsScalar (stats_avg [1, 12]) + ", Max = " + castAsScalar (stats_max [1, 12]));
- print (" True Negatives: Min = " + castAsScalar (stats_min [1, 13]) + ", Avg = " + castAsScalar (stats_avg [1, 13]) + ", Max = " + castAsScalar (stats_max [1, 13]));
- print (" False Negatives: Min = " + castAsScalar (stats_min [1, 14]) + ", Avg = " + castAsScalar (stats_avg [1, 14]) + ", Max = " + castAsScalar (stats_max [1, 14]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 15]) + ", Avg = " + castAsScalar (stats_avg [1, 15]) + ", Max = " + castAsScalar (stats_max [1, 15]));
- print ("Recall (Sensit-y)%: Min = " + castAsScalar (stats_min [1, 16]) + ", Avg = " + castAsScalar (stats_avg [1, 16]) + ", Max = " + castAsScalar (stats_max [1, 16]));
- print (" Specificity %: Min = " + castAsScalar (stats_min [1, 17]) + ", Avg = " + castAsScalar (stats_avg [1, 17]) + ", Max = " + castAsScalar (stats_max [1, 17]));
- print (" Value - Cost: Min = " + castAsScalar (stats_min [1, 18]) + ", Avg = " + castAsScalar (stats_avg [1, 18]) + ", Max = " + castAsScalar (stats_max [1, 18]));
+ print (" True Positives: Min = " + as.scalar (stats_min [1, 11]) + ", Avg = " + as.scalar (stats_avg [1, 11]) + ", Max = " + as.scalar (stats_max [1, 11]));
+ print (" False Positives: Min = " + as.scalar (stats_min [1, 12]) + ", Avg = " + as.scalar (stats_avg [1, 12]) + ", Max = " + as.scalar (stats_max [1, 12]));
+ print (" True Negatives: Min = " + as.scalar (stats_min [1, 13]) + ", Avg = " + as.scalar (stats_avg [1, 13]) + ", Max = " + as.scalar (stats_max [1, 13]));
+ print (" False Negatives: Min = " + as.scalar (stats_min [1, 14]) + ", Avg = " + as.scalar (stats_avg [1, 14]) + ", Max = " + as.scalar (stats_max [1, 14]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 15]) + ", Avg = " + as.scalar (stats_avg [1, 15]) + ", Max = " + as.scalar (stats_max [1, 15]));
+ print ("Recall (Sensit-y)%: Min = " + as.scalar (stats_min [1, 16]) + ", Avg = " + as.scalar (stats_avg [1, 16]) + ", Max = " + as.scalar (stats_max [1, 16]));
+ print (" Specificity %: Min = " + as.scalar (stats_min [1, 17]) + ", Avg = " + as.scalar (stats_avg [1, 17]) + ", Max = " + as.scalar (stats_max [1, 17]));
+ print (" Value - Cost: Min = " + as.scalar (stats_min [1, 18]) + ", Avg = " + as.scalar (stats_avg [1, 18]) + ", Max = " + as.scalar (stats_max [1, 18]));
print (" ");
if (1==1) {
print(" ")
}
/*
print ("TEST Data, Model-estimated statistics:");
- print (" True Positives: Min = " + castAsScalar (stats_min [1, 21]) + ", Avg = " + castAsScalar (stats_avg [1, 21]) + ", Max = " + castAsScalar (stats_max [1, 21]));
- print (" False Positives: Min = " + castAsScalar (stats_min [1, 22]) + ", Avg = " + castAsScalar (stats_avg [1, 22]) + ", Max = " + castAsScalar (stats_max [1, 22]));
- print (" True Negatives: Min = " + castAsScalar (stats_min [1, 23]) + ", Avg = " + castAsScalar (stats_avg [1, 23]) + ", Max = " + castAsScalar (stats_max [1, 23]));
- print (" False Negatives: Min = " + castAsScalar (stats_min [1, 24]) + ", Avg = " + castAsScalar (stats_avg [1, 24]) + ", Max = " + castAsScalar (stats_max [1, 24]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 25]) + ", Avg = " + castAsScalar (stats_avg [1, 25]) + ", Max = " + castAsScalar (stats_max [1, 25]));
- print ("Recall (Sensit-y)%: Min = " + castAsScalar (stats_min [1, 26]) + ", Avg = " + castAsScalar (stats_avg [1, 26]) + ", Max = " + castAsScalar (stats_max [1, 26]));
- print (" Specificity %: Min = " + castAsScalar (stats_min [1, 27]) + ", Avg = " + castAsScalar (stats_avg [1, 27]) + ", Max = " + castAsScalar (stats_max [1, 27]));
- print (" Value - Cost: Min = " + castAsScalar (stats_min [1, 28]) + ", Avg = " + castAsScalar (stats_avg [1, 28]) + ", Max = " + castAsScalar (stats_max [1, 28]));
+ print (" True Positives: Min = " + as.scalar (stats_min [1, 21]) + ", Avg = " + as.scalar (stats_avg [1, 21]) + ", Max = " + as.scalar (stats_max [1, 21]));
+ print (" False Positives: Min = " + as.scalar (stats_min [1, 22]) + ", Avg = " + as.scalar (stats_avg [1, 22]) + ", Max = " + as.scalar (stats_max [1, 22]));
+ print (" True Negatives: Min = " + as.scalar (stats_min [1, 23]) + ", Avg = " + as.scalar (stats_avg [1, 23]) + ", Max = " + as.scalar (stats_max [1, 23]));
+ print (" False Negatives: Min = " + as.scalar (stats_min [1, 24]) + ", Avg = " + as.scalar (stats_avg [1, 24]) + ", Max = " + as.scalar (stats_max [1, 24]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 25]) + ", Avg = " + as.scalar (stats_avg [1, 25]) + ", Max = " + as.scalar (stats_max [1, 25]));
+ print ("Recall (Sensit-y)%: Min = " + as.scalar (stats_min [1, 26]) + ", Avg = " + as.scalar (stats_avg [1, 26]) + ", Max = " + as.scalar (stats_max [1, 26]));
+ print (" Specificity %: Min = " + as.scalar (stats_min [1, 27]) + ", Avg = " + as.scalar (stats_avg [1, 27]) + ", Max = " + as.scalar (stats_max [1, 27]));
+ print (" Value - Cost: Min = " + as.scalar (stats_min [1, 28]) + ", Avg = " + as.scalar (stats_avg [1, 28]) + ", Max = " + as.scalar (stats_max [1, 28]));
print (" ");
if (1==1) {
print(" ")
}
*/
print ("TEST Data, Label comparison statistics:");
- print (" True Positives: Min = " + castAsScalar (stats_min [1, 31]) + ", Avg = " + castAsScalar (stats_avg [1, 31]) + ", Max = " + castAsScalar (stats_max [1, 31]));
- print (" False Positives: Min = " + castAsScalar (stats_min [1, 32]) + ", Avg = " + castAsScalar (stats_avg [1, 32]) + ", Max = " + castAsScalar (stats_max [1, 32]));
- print (" True Negatives: Min = " + castAsScalar (stats_min [1, 33]) + ", Avg = " + castAsScalar (stats_avg [1, 33]) + ", Max = " + castAsScalar (stats_max [1, 33]));
- print (" False Negatives: Min = " + castAsScalar (stats_min [1, 34]) + ", Avg = " + castAsScalar (stats_avg [1, 34]) + ", Max = " + castAsScalar (stats_max [1, 34]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 35]) + ", Avg = " + castAsScalar (stats_avg [1, 35]) + ", Max = " + castAsScalar (stats_max [1, 35]));
- print ("Recall (Sensit-y)%: Min = " + castAsScalar (stats_min [1, 36]) + ", Avg = " + castAsScalar (stats_avg [1, 36]) + ", Max = " + castAsScalar (stats_max [1, 36]));
- print (" Specificity %: Min = " + castAsScalar (stats_min [1, 37]) + ", Avg = " + castAsScalar (stats_avg [1, 37]) + ", Max = " + castAsScalar (stats_max [1, 37]));
- print (" Value - Cost: Min = " + castAsScalar (stats_min [1, 38]) + ", Avg = " + castAsScalar (stats_avg [1, 38]) + ", Max = " + castAsScalar (stats_max [1, 38]));
+ print (" True Positives: Min = " + as.scalar (stats_min [1, 31]) + ", Avg = " + as.scalar (stats_avg [1, 31]) + ", Max = " + as.scalar (stats_max [1, 31]));
+ print (" False Positives: Min = " + as.scalar (stats_min [1, 32]) + ", Avg = " + as.scalar (stats_avg [1, 32]) + ", Max = " + as.scalar (stats_max [1, 32]));
+ print (" True Negatives: Min = " + as.scalar (stats_min [1, 33]) + ", Avg = " + as.scalar (stats_avg [1, 33]) + ", Max = " + as.scalar (stats_max [1, 33]));
+ print (" False Negatives: Min = " + as.scalar (stats_min [1, 34]) + ", Avg = " + as.scalar (stats_avg [1, 34]) + ", Max = " + as.scalar (stats_max [1, 34]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 35]) + ", Avg = " + as.scalar (stats_avg [1, 35]) + ", Max = " + as.scalar (stats_max [1, 35]));
+ print ("Recall (Sensit-y)%: Min = " + as.scalar (stats_min [1, 36]) + ", Avg = " + as.scalar (stats_avg [1, 36]) + ", Max = " + as.scalar (stats_max [1, 36]));
+ print (" Specificity %: Min = " + as.scalar (stats_min [1, 37]) + ", Avg = " + as.scalar (stats_avg [1, 37]) + ", Max = " + as.scalar (stats_max [1, 37]));
+ print (" Value - Cost: Min = " + as.scalar (stats_min [1, 38]) + ", Avg = " + as.scalar (stats_avg [1, 38]) + ", Max = " + as.scalar (stats_max [1, 38]));
err = 0;
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/validation/CV_MultiClassSVM.sasha.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/validation/CV_MultiClassSVM.sasha.dml b/src/test/scripts/applications/validation/CV_MultiClassSVM.sasha.dml
index f63c636..a3099d6 100644
--- a/src/test/scripts/applications/validation/CV_MultiClassSVM.sasha.dml
+++ b/src/test/scripts/applications/validation/CV_MultiClassSVM.sasha.dml
@@ -309,18 +309,18 @@ printFoldStatistics = function (Matrix[double] stats ) return( Integer err )
stats_min = round (colMins (stats) * 10000.0) / 10000.0;
print ("Training Data, Label comparison statistics:");
- print (" True Matches: Min = " + castAsScalar (stats_min [1, 11]) + ", Avg = " + castAsScalar (stats_avg [1, 11]) + ", Max = " + castAsScalar (stats_max [1, 11]));
- print (" False Matches: Min = " + castAsScalar (stats_min [1, 12]) + ", Avg = " + castAsScalar (stats_avg [1, 12]) + ", Max = " + castAsScalar (stats_max [1, 12]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 15]) + ", Avg = " + castAsScalar (stats_avg [1, 15]) + ", Max = " + castAsScalar (stats_max [1, 15]));
+ print (" True Matches: Min = " + as.scalar (stats_min [1, 11]) + ", Avg = " + as.scalar (stats_avg [1, 11]) + ", Max = " + as.scalar (stats_max [1, 11]));
+ print (" False Matches: Min = " + as.scalar (stats_min [1, 12]) + ", Avg = " + as.scalar (stats_avg [1, 12]) + ", Max = " + as.scalar (stats_max [1, 12]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 15]) + ", Avg = " + as.scalar (stats_avg [1, 15]) + ", Max = " + as.scalar (stats_max [1, 15]));
print (" ");
if (1==1) {
print(" ")
}
print ("TEST Data, Label comparison statistics:");
- print (" True Matches: Min = " + castAsScalar (stats_min [1, 31]) + ", Avg = " + castAsScalar (stats_avg [1, 31]) + ", Max = " + castAsScalar (stats_max [1, 31]));
- print (" False Matches: Min = " + castAsScalar (stats_min [1, 32]) + ", Avg = " + castAsScalar (stats_avg [1, 32]) + ", Max = " + castAsScalar (stats_max [1, 32]));
- print (" Precision %: Min = " + castAsScalar (stats_min [1, 35]) + ", Avg = " + castAsScalar (stats_avg [1, 35]) + ", Max = " + castAsScalar (stats_max [1, 35]));
+ print (" True Matches: Min = " + as.scalar (stats_min [1, 31]) + ", Avg = " + as.scalar (stats_avg [1, 31]) + ", Max = " + as.scalar (stats_max [1, 31]));
+ print (" False Matches: Min = " + as.scalar (stats_min [1, 32]) + ", Avg = " + as.scalar (stats_avg [1, 32]) + ", Max = " + as.scalar (stats_max [1, 32]));
+ print (" Precision %: Min = " + as.scalar (stats_min [1, 35]) + ", Avg = " + as.scalar (stats_avg [1, 35]) + ", Max = " + as.scalar (stats_max [1, 35]));
err = 0;
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/validation/LinearLogisticRegression.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/validation/LinearLogisticRegression.dml b/src/test/scripts/applications/validation/LinearLogisticRegression.dml
index 473b2dc..1444bd4 100644
--- a/src/test/scripts/applications/validation/LinearLogisticRegression.dml
+++ b/src/test/scripts/applications/validation/LinearLogisticRegression.dml
@@ -105,7 +105,7 @@ while(!converge) {
norm_grad = sqrt(sum(grad*grad))
print("-- Outer Iteration = " + iter)
- objScalar = castAsScalar(obj)
+ objScalar = as.scalar(obj)
print(" Iterations = " + iter + ", Objective = " + objScalar + ", Gradient Norm = " + norm_grad)
# SOLVE TRUST REGION SUB-PROBLEM
@@ -123,23 +123,23 @@ while(!converge) {
alpha_deno = t(d) %*% Hd
alpha = norm_r2 / alpha_deno
- s = s + castAsScalar(alpha) * d
- os = os + castAsScalar(alpha) * od
+ s = s + as.scalar(alpha) * d
+ os = os + as.scalar(alpha) * od
sts = t(s) %*% s
delta2 = delta*delta
- stsScalar = castAsScalar(sts)
+ stsScalar = as.scalar(sts)
shouldBreak = FALSE; # to mimic "break" in the following 'if' condition
if (stsScalar > delta2) {
print(" --- cg reaches trust region boundary")
- s = s - castAsScalar(alpha) * d
- os = os - castAsScalar(alpha) * od
+ s = s - as.scalar(alpha) * d
+ os = os - as.scalar(alpha) * od
std = t(s) %*% d
dtd = t(d) %*% d
sts = t(s) %*% s
rad = sqrt(std*std + dtd*(delta2 - sts))
- stdScalar = castAsScalar(std)
+ stdScalar = as.scalar(std)
if(stdScalar >= 0) {
tau = (delta2 - sts)/(std + rad)
}
@@ -147,9 +147,9 @@ while(!converge) {
tau = (rad - std)/dtd
}
- s = s + castAsScalar(tau) * d
- os = os + castAsScalar(tau) * od
- r = r - castAsScalar(tau) * Hd
+ s = s + as.scalar(tau) * d
+ os = os + as.scalar(tau) * od
+ r = r - as.scalar(tau) * Hd
#break
shouldBreak = TRUE;
@@ -158,7 +158,7 @@ while(!converge) {
}
if (!shouldBreak) {
- r = r - castAsScalar(alpha) * Hd
+ r = r - as.scalar(alpha) * Hd
old_norm_r2 = norm_r2
norm_r2 = sum(r*r)
beta = norm_r2/old_norm_r2
@@ -179,10 +179,10 @@ while(!converge) {
objnew = 0.5 * t(wnew) %*% wnew + C * sum(-log(logisticnew))
actred = (obj - objnew)
- actredScalar = castAsScalar(actred)
+ actredScalar = as.scalar(actred)
rho = actred / qk
- qkScalar = castAsScalar(qk)
- rhoScalar = castAsScalar(rho);
+ qkScalar = as.scalar(qk)
+ rhoScalar = as.scalar(rho);
snorm = sqrt(sum( s * s ))
print(" Actual = " + actredScalar)
@@ -192,12 +192,12 @@ while(!converge) {
delta = min(delta, snorm)
}
alpha2 = objnew - obj - gs
- alpha2Scalar = castAsScalar(alpha2)
+ alpha2Scalar = as.scalar(alpha2)
if (alpha2Scalar <= 0) {
alpha = sigma3*e
}
else {
- ascalar = max(sigma1, -0.5*castAsScalar(gs)/alpha2Scalar)
+ ascalar = max(sigma1, -0.5*as.scalar(gs)/alpha2Scalar)
alpha = ascalar*e
}
@@ -211,7 +211,7 @@ while(!converge) {
obj = objnew
}
- alphaScalar = castAsScalar(alpha)
+ alphaScalar = as.scalar(alpha)
if (rhoScalar < eta0){
delta = min(max( alphaScalar , sigma1) * snorm, sigma2 * delta )
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/validation/genRandData4LogisticRegression.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/validation/genRandData4LogisticRegression.dml b/src/test/scripts/applications/validation/genRandData4LogisticRegression.dml
index b42a315..c7cd4a2 100644
--- a/src/test/scripts/applications/validation/genRandData4LogisticRegression.dml
+++ b/src/test/scripts/applications/validation/genRandData4LogisticRegression.dml
@@ -101,17 +101,17 @@ scaleWeights =
W_ext [, 1] = w_unscaled;
S1 = colSums (X_data %*% W_ext);
TF = Rand (rows = 2, cols = 2, min = 1, max = 1);
- TF [1, 1] = S1 [1, 1] * meanLF * nrow (X_data) / castAsScalar (S1 %*% t(S1));
+ TF [1, 1] = S1 [1, 1] * meanLF * nrow (X_data) / as.scalar (S1 %*% t(S1));
TF [1, 2] = S1 [1, 2];
- TF [2, 1] = S1 [1, 2] * meanLF * nrow (X_data) / castAsScalar (S1 %*% t(S1));
+ TF [2, 1] = S1 [1, 2] * meanLF * nrow (X_data) / as.scalar (S1 %*% t(S1));
TF [2, 2] = - S1 [1, 1];
TF = W_ext %*% TF;
Q = t(TF) %*% t(X_data) %*% X_data %*% TF;
Q [1, 1] = Q [1, 1] - nrow (X_data) * meanLF * meanLF;
new_sigmaLF = sigmaLF;
- discr = castAsScalar (Q [1, 1] * Q [2, 2] - Q [1, 2] * Q [2, 1] - nrow (X_data) * Q [2, 2] * sigmaLF * sigmaLF);
+ discr = as.scalar (Q [1, 1] * Q [2, 2] - Q [1, 2] * Q [2, 1] - nrow (X_data) * Q [2, 2] * sigmaLF * sigmaLF);
if (discr > 0.0) {
- new_sigmaLF = sqrt (castAsScalar ((Q [1, 1] * Q [2, 2] - Q [1, 2] * Q [2, 1]) / (nrow (X_data) * Q [2, 2])));
+ new_sigmaLF = sqrt (as.scalar ((Q [1, 1] * Q [2, 2] - Q [1, 2] * Q [2, 1]) / (nrow (X_data) * Q [2, 2])));
discr = -0.0;
}
t = Rand (rows = 2, cols = 1, min = 1, max = 1);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/gdfo/LinregCG.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/gdfo/LinregCG.dml b/src/test/scripts/functions/gdfo/LinregCG.dml
index 85a66e4..92f15d7 100644
--- a/src/test/scripts/functions/gdfo/LinregCG.dml
+++ b/src/test/scripts/functions/gdfo/LinregCG.dml
@@ -39,7 +39,7 @@ w = matrix(0, rows = ncol(X), cols = 1);
i = 0;
while(i < maxiter) {
q = ((t(X) %*% (X %*% p)) + eps * p);
- alpha = norm_r2 / castAsScalar(t(p) %*% q);
+ alpha = norm_r2 / as.scalar(t(p) %*% q);
w = w + alpha * p;
old_norm_r2 = norm_r2;
r = r + alpha * q;
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/jmlc/reuse-glm-predict.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/jmlc/reuse-glm-predict.dml b/src/test/scripts/functions/jmlc/reuse-glm-predict.dml
index 9a29d67..6ff0b68 100644
--- a/src/test/scripts/functions/jmlc/reuse-glm-predict.dml
+++ b/src/test/scripts/functions/jmlc/reuse-glm-predict.dml
@@ -251,15 +251,15 @@ if (fileY != " ")
str = append (str, "DEVIANCE_G2_PVAL,,TRUE," + G2_scaled_pValue);
for (i in 1:ncol(Y)) {
- str = append (str, "AVG_TOT_Y," + i + ",," + castAsScalar (avg_tot_Y [1, i]));
- str = append (str, "STDEV_TOT_Y," + i + ",," + castAsScalar (sqrt (var_tot_Y [1, i])));
- str = append (str, "AVG_RES_Y," + i + ",," + castAsScalar (avg_res_Y [1, i]));
- str = append (str, "STDEV_RES_Y," + i + ",," + castAsScalar (sqrt (var_res_Y [1, i])));
- str = append (str, "PRED_STDEV_RES," + i + ",TRUE," + castAsScalar (sqrt (predicted_avg_var_res_Y [1, i])));
- str = append (str, "PLAIN_R2," + i + ",," + castAsScalar (plain_R2 [1, i]));
- str = append (str, "ADJUSTED_R2," + i + ",," + castAsScalar (adjust_R2 [1, i]));
- str = append (str, "PLAIN_R2_NOBIAS," + i + ",," + castAsScalar (plain_R2_nobias [1, i]));
- str = append (str, "ADJUSTED_R2_NOBIAS," + i + ",," + castAsScalar (adjust_R2_nobias [1, i]));
+ str = append (str, "AVG_TOT_Y," + i + ",," + as.scalar (avg_tot_Y [1, i]));
+ str = append (str, "STDEV_TOT_Y," + i + ",," + as.scalar (sqrt (var_tot_Y [1, i])));
+ str = append (str, "AVG_RES_Y," + i + ",," + as.scalar (avg_res_Y [1, i]));
+ str = append (str, "STDEV_RES_Y," + i + ",," + as.scalar (sqrt (var_res_Y [1, i])));
+ str = append (str, "PRED_STDEV_RES," + i + ",TRUE," + as.scalar (sqrt (predicted_avg_var_res_Y [1, i])));
+ str = append (str, "PLAIN_R2," + i + ",," + as.scalar (plain_R2 [1, i]));
+ str = append (str, "ADJUSTED_R2," + i + ",," + as.scalar (adjust_R2 [1, i]));
+ str = append (str, "PLAIN_R2_NOBIAS," + i + ",," + as.scalar (plain_R2_nobias [1, i]));
+ str = append (str, "ADJUSTED_R2_NOBIAS," + i + ",," + as.scalar (adjust_R2_nobias [1, i]));
}
if (fileO != " ") {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/misc/IPAUnknownRecursion.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/misc/IPAUnknownRecursion.dml b/src/test/scripts/functions/misc/IPAUnknownRecursion.dml
index 22570b8..f6be7b0 100644
--- a/src/test/scripts/functions/misc/IPAUnknownRecursion.dml
+++ b/src/test/scripts/functions/misc/IPAUnknownRecursion.dml
@@ -28,7 +28,7 @@ factorial = function(Matrix[Double] arr, Integer pos) return (Matrix[Double] arr
}
for(i in 1:ncol(arr))
- print("inside factorial (" + pos + ") " + i + ": " + castAsScalar(arr[1, i]))
+ print("inside factorial (" + pos + ") " + i + ": " + as.scalar(arr[1, i]))
}
n = $1
@@ -38,7 +38,7 @@ arr = factorial(arr, n)
R = matrix(0, rows=1, cols=n);
for(i in 1:n) #copy important to test dynamic rewrites
{
- print("main factorial " + i + ": " + castAsScalar(arr[1, i]))
+ print("main factorial " + i + ": " + as.scalar(arr[1, i]))
R[1,i] = as.scalar(arr[1, i]);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/misc/dt_change_4b.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/misc/dt_change_4b.dml b/src/test/scripts/functions/misc/dt_change_4b.dml
index c5d8c7e..d443587 100644
--- a/src/test/scripts/functions/misc/dt_change_4b.dml
+++ b/src/test/scripts/functions/misc/dt_change_4b.dml
@@ -22,7 +22,7 @@
Y = matrix(1, rows=10, cols=10);
X = matrix(7, rows=10, cols=10);
-X = castAsScalar(X[1,1]);
+X = as.scalar(X[1,1]);
print("Result: "+sum(X + Y));
#expected: "Result: 800.0"
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/misc/dt_change_4c.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/misc/dt_change_4c.dml b/src/test/scripts/functions/misc/dt_change_4c.dml
index ba561bf..3cf60fa 100644
--- a/src/test/scripts/functions/misc/dt_change_4c.dml
+++ b/src/test/scripts/functions/misc/dt_change_4c.dml
@@ -24,7 +24,7 @@ foo = function(Matrix[Double] input) return (Double out)
{
if( 1==1 ){} #prevent inlining
- out = castAsScalar(input[1,1]);
+ out = as.scalar(input[1,1]);
}
Y = matrix(1, rows=10, cols=10);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/misc/dt_change_4f.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/misc/dt_change_4f.dml b/src/test/scripts/functions/misc/dt_change_4f.dml
index 71f083c..1050a2c 100644
--- a/src/test/scripts/functions/misc/dt_change_4f.dml
+++ b/src/test/scripts/functions/misc/dt_change_4f.dml
@@ -23,7 +23,7 @@
Y = matrix(1, rows=10, cols=10);
X = matrix(7, rows=10, cols=10);
if(1==1){}
-X = castAsScalar(X[1,1]);
+X = as.scalar(X[1,1]);
print("Result: "+sum(X + Y));
#expected: "Result: 800.0"
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor35.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor35.dml b/src/test/scripts/functions/parfor/parfor35.dml
index 81f7564..b1c8da9 100644
--- a/src/test/scripts/functions/parfor/parfor35.dml
+++ b/src/test/scripts/functions/parfor/parfor35.dml
@@ -26,7 +26,7 @@ dummy = matrix(1, rows=1,cols=1);
parfor( i in 1:20 )
{
- val = castAsScalar(B[i,i]);
+ val = as.scalar(B[i,i]);
b = A[i,val]; #due to parser change A[i,B[i,]];
c = dummy*(b+i);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor48b.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor48b.dml b/src/test/scripts/functions/parfor/parfor48b.dml
index c87f920..edeeea4 100644
--- a/src/test/scripts/functions/parfor/parfor48b.dml
+++ b/src/test/scripts/functions/parfor/parfor48b.dml
@@ -22,7 +22,7 @@
A = Rand(rows=10, cols=10, min=0.0, max=1.0, sparsity=1.0)
-parfor(i in 1:castAsScalar(A[1,1])){
+parfor(i in 1:as.scalar(A[1,1])){
parfor(j in 1:A+ncol(A)){
print("i="+i+", j="+j);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor6.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor6.dml b/src/test/scripts/functions/parfor/parfor6.dml
index aae1370..e50b892 100644
--- a/src/test/scripts/functions/parfor/parfor6.dml
+++ b/src/test/scripts/functions/parfor/parfor6.dml
@@ -24,6 +24,6 @@ A = Rand(rows=10,cols=1);
parfor( i in 1:10 )
{
- b = i + castAsScalar(A[i,1]);
+ b = i + as.scalar(A[i,1]);
#print(b);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor7.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor7.dml b/src/test/scripts/functions/parfor/parfor7.dml
index 39c14ad..807239b 100644
--- a/src/test/scripts/functions/parfor/parfor7.dml
+++ b/src/test/scripts/functions/parfor/parfor7.dml
@@ -24,7 +24,7 @@ A = Rand(rows=10,cols=1);
parfor( i in 2:10 )
{
- b = i + castAsScalar(A[i,1]) + castAsScalar(A[i+1,1]);
+ b = i + as.scalar(A[i,1]) + as.scalar(A[i+1,1]);
#print(b);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor8.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor8.dml b/src/test/scripts/functions/parfor/parfor8.dml
index 21d96ee..73f1e25 100644
--- a/src/test/scripts/functions/parfor/parfor8.dml
+++ b/src/test/scripts/functions/parfor/parfor8.dml
@@ -25,7 +25,7 @@ a = 1
parfor( i in 2:10 )
{
- b = a + castAsScalar(A[i,1]) + castAsScalar(A[i+1,1]);
+ b = a + as.scalar(A[i,1]) + as.scalar(A[i+1,1]);
a = i;
# print(b);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor9.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor9.dml b/src/test/scripts/functions/parfor/parfor9.dml
index fb50cb7..b9b970f 100644
--- a/src/test/scripts/functions/parfor/parfor9.dml
+++ b/src/test/scripts/functions/parfor/parfor9.dml
@@ -25,7 +25,7 @@ a = 1
parfor( i in 2:10 )
{
- b = a + castAsScalar(A[i,1]) + castAsScalar(A[i-1,1]);
+ b = a + as.scalar(A[i,1]) + as.scalar(A[i-1,1]);
a = i;
#print(b);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor_optimizer2.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor_optimizer2.dml b/src/test/scripts/functions/parfor/parfor_optimizer2.dml
index e6007af..bc8cdc0 100644
--- a/src/test/scripts/functions/parfor/parfor_optimizer2.dml
+++ b/src/test/scripts/functions/parfor/parfor_optimizer2.dml
@@ -73,14 +73,14 @@ dummy = matrix(1, rows=1, cols=1);
parfor( i in 1:s1size, check=0, opt=RULEBASED) {
- a1 = castAsScalar(S1[,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, check=0) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
if (k1 == k2) {
@@ -233,7 +233,7 @@ computeRanks = function(Matrix[Double] X) return (Matrix[Double] Ranks) {
if( i>1 ){
prefixSum = sum(X[1:(i-1),1]);
}
- Rks[i,1] = dummy * (prefixSum + ((castAsScalar(X[i,1])+1)/2));
+ Rks[i,1] = dummy * (prefixSum + ((as.scalar(X[i,1])+1)/2));
}
Ranks = Rks;
}
@@ -268,7 +268,7 @@ bivar_oo = function(Matrix[Double] A, Matrix[Double] B) return (Double sp) {
covXY = 0.0;
for(i in 1:catA) {
- covXY = covXY + sum((F[i,]/(W-1)) * (castAsScalar(C[i,1])-meanX) * (t(D[,1])-meanY));
+ covXY = covXY + sum((F[i,]/(W-1)) * (as.scalar(C[i,1])-meanX) * (t(D[,1])-meanY));
}
sp = covXY/(sqrt(varX)*sqrt(varY));
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor_threadid_recompile1.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor_threadid_recompile1.dml b/src/test/scripts/functions/parfor/parfor_threadid_recompile1.dml
index 1995f9d..0370865 100644
--- a/src/test/scripts/functions/parfor/parfor_threadid_recompile1.dml
+++ b/src/test/scripts/functions/parfor/parfor_threadid_recompile1.dml
@@ -29,17 +29,17 @@ bin_defns = matrix(0, rows=num_bin_defns, cols=2)
attr2pos = matrix(0, rows=nrow(A), cols=2)
pos = 1
for(i in 1:nrow(A)){
- number_of_bins = castAsScalar(A[i,1])
+ number_of_bins = as.scalar(A[i,1])
attr2pos[i,1] = pos
attr2pos[i,2] = pos + number_of_bins - 1
pos = pos + number_of_bins
}
for(i in 1:nrow(A), check=0){
- num_bins = castAsScalar(A[i,1])
+ num_bins = as.scalar(A[i,1])
- start_position = castAsScalar(attr2pos[i,1])
- end_position = castAsScalar(attr2pos[i,2])
+ start_position = as.scalar(attr2pos[i,1])
+ end_position = as.scalar(attr2pos[i,2])
#SEQ CALL 1
bin_defns[start_position:end_position,1] = seq(1, num_bins, 1)
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/parfor/parfor_threadid_recompile2.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/parfor/parfor_threadid_recompile2.dml b/src/test/scripts/functions/parfor/parfor_threadid_recompile2.dml
index ab89580..7ce8360 100644
--- a/src/test/scripts/functions/parfor/parfor_threadid_recompile2.dml
+++ b/src/test/scripts/functions/parfor/parfor_threadid_recompile2.dml
@@ -29,17 +29,17 @@ bin_defns = matrix(0, rows=num_bin_defns, cols=2)
attr2pos = matrix(0, rows=nrow(A), cols=2)
pos = 1
for(i in 1:nrow(A)){
- number_of_bins = castAsScalar(A[i,1])
+ number_of_bins = as.scalar(A[i,1])
attr2pos[i,1] = pos
attr2pos[i,2] = pos + number_of_bins - 1
pos = pos + number_of_bins
}
parfor(i in 1:nrow(A), check=0){
- num_bins = castAsScalar(A[i,1])
+ num_bins = as.scalar(A[i,1])
- start_position = castAsScalar(attr2pos[i,1])
- end_position = castAsScalar(attr2pos[i,2])
+ start_position = as.scalar(attr2pos[i,1])
+ end_position = as.scalar(attr2pos[i,2])
#SEQ CALL 1
bin_defns[start_position:end_position,1] = seq(1, num_bins, 1)
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/recompile/for_recompile.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/recompile/for_recompile.dml b/src/test/scripts/functions/recompile/for_recompile.dml
index 96d7dac..8947b99 100644
--- a/src/test/scripts/functions/recompile/for_recompile.dml
+++ b/src/test/scripts/functions/recompile/for_recompile.dml
@@ -22,7 +22,7 @@
V = Rand(rows=$1+1, cols=$2+1, min=$3, max=$3);
Z = Rand(rows=1,cols=1,min=0,max=0);
-for( i in $3:castAsScalar(V[1,1]) )
+for( i in $3:as.scalar(V[1,1]) )
{
Z[1,1] = V[1,1];
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/recompile/if_recompile.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/recompile/if_recompile.dml b/src/test/scripts/functions/recompile/if_recompile.dml
index 2d02e01..91f435b 100644
--- a/src/test/scripts/functions/recompile/if_recompile.dml
+++ b/src/test/scripts/functions/recompile/if_recompile.dml
@@ -22,7 +22,7 @@
V = Rand(rows=$1+1, cols=$2+1, min=$3, max=$3);
Z = Rand(rows=1,cols=1,min=0,max=0);
-if( castAsScalar(V[1,1])>castAsScalar(Z[1,1]) )
+if( as.scalar(V[1,1])>as.scalar(Z[1,1]) )
{
Z[1,1] = V[1,1];
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/recompile/parfor_recompile.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/recompile/parfor_recompile.dml b/src/test/scripts/functions/recompile/parfor_recompile.dml
index 5e14440..f8223e0 100644
--- a/src/test/scripts/functions/recompile/parfor_recompile.dml
+++ b/src/test/scripts/functions/recompile/parfor_recompile.dml
@@ -22,7 +22,7 @@
V = Rand(rows=$1+1, cols=$2+1, min=$3, max=$3);
Z = Rand(rows=1,cols=1,min=0,max=0);
-parfor( i in $3:castAsScalar(V[1,1]), check=0 )
+parfor( i in $3:as.scalar(V[1,1]), check=0 )
{
Z[1,1] = V[1,1];
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/recompile/while_recompile.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/recompile/while_recompile.dml b/src/test/scripts/functions/recompile/while_recompile.dml
index 05dd424..de74b0d 100644
--- a/src/test/scripts/functions/recompile/while_recompile.dml
+++ b/src/test/scripts/functions/recompile/while_recompile.dml
@@ -22,7 +22,7 @@
V = Rand(rows=$1+1, cols=$2+1, min=$3, max=$3);
Z = Rand(rows=1,cols=1,min=0,max=0);
-while( castAsScalar(V[1,1])>castAsScalar(Z[1,1]) )
+while( as.scalar(V[1,1])>as.scalar(Z[1,1]) )
{
Z[1,1] = V[1,1];
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/unary/matrix/CastAsScalarTest.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/unary/matrix/CastAsScalarTest.dml b/src/test/scripts/functions/unary/matrix/CastAsScalarTest.dml
index b566e44..7731cac 100644
--- a/src/test/scripts/functions/unary/matrix/CastAsScalarTest.dml
+++ b/src/test/scripts/functions/unary/matrix/CastAsScalarTest.dml
@@ -24,6 +24,6 @@
$$readhelper$$
A = read("$$indir$$a", rows=1, cols=1, format="text");
-b = castAsScalar(A);
+b = as.scalar(A);
BHelper = b * Helper;
write(BHelper, "$$outdir$$b", format="text");
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/unary/matrix/eigen.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/unary/matrix/eigen.dml b/src/test/scripts/functions/unary/matrix/eigen.dml
index e4c7c46..12a2f87 100644
--- a/src/test/scripts/functions/unary/matrix/eigen.dml
+++ b/src/test/scripts/functions/unary/matrix/eigen.dml
@@ -40,7 +40,7 @@ numEval = $2;
D = matrix(1, numEval, 1);
for ( i in 1:numEval ) {
Av = A %*% evec[,i];
- rhs = castAsScalar(eval[i,1]) * evec[,i];
+ rhs = as.scalar(eval[i,1]) * evec[,i];
diff = sum(Av-rhs);
D[i,1] = diff;
}
@@ -49,7 +49,7 @@ for ( i in 1:numEval ) {
# TODO: dummy if() must be removed
v = evec[,1];
Av = A %*% v;
-rhs = castAsScalar(eval[1,1]) * evec[,1];
+rhs = as.scalar(eval[1,1]) * evec[,1];
diff = sum(Av-rhs);
D = matrix(1,1,1);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/functions/unary/matrix/qr.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/functions/unary/matrix/qr.dml b/src/test/scripts/functions/unary/matrix/qr.dml
index 184ff80..5454685 100644
--- a/src/test/scripts/functions/unary/matrix/qr.dml
+++ b/src/test/scripts/functions/unary/matrix/qr.dml
@@ -38,7 +38,7 @@ eye = diag(ones);
Q = eye;
for( j in n:1 ) {
v = H[,j];
- Qj = eye - 2 * (v %*% t(v))/castAsScalar((t(v)%*%v));
+ Qj = eye - 2 * (v %*% t(v))/as.scalar((t(v)%*%v));
Q = Qj %*% Q;
}
[2/2] incubator-systemml git commit: [SYSTEMML-647] Replace
castAsScalar calls
Posted by de...@apache.org.
[SYSTEMML-647] Replace castAsScalar calls
Replace castAsScalar() with as.scalar() in DML.
Replace castAsScalar() with scalar() in PYDML.
Closes #136.
Project: http://git-wip-us.apache.org/repos/asf/incubator-systemml/repo
Commit: http://git-wip-us.apache.org/repos/asf/incubator-systemml/commit/2da81457
Tree: http://git-wip-us.apache.org/repos/asf/incubator-systemml/tree/2da81457
Diff: http://git-wip-us.apache.org/repos/asf/incubator-systemml/diff/2da81457
Branch: refs/heads/master
Commit: 2da8145744f52d725e56e883997ef9ad7206a9b3
Parents: 7013910
Author: Deron Eriksson <de...@us.ibm.com>
Authored: Wed May 4 18:10:24 2016 -0700
Committer: Deron Eriksson <de...@us.ibm.com>
Committed: Wed May 4 18:11:05 2016 -0700
----------------------------------------------------------------------
scripts/algorithms/GLM-predict.dml | 18 ++--
scripts/algorithms/GLM.dml | 6 +-
scripts/algorithms/Kmeans-predict.dml | 36 +++----
scripts/algorithms/Kmeans.dml | 8 +-
scripts/algorithms/StepGLM.dml | 6 +-
scripts/algorithms/Univar-Stats.dml | 4 +-
scripts/algorithms/bivar-stats.dml | 20 ++--
scripts/algorithms/l2-svm-predict.dml | 2 +-
scripts/algorithms/m-svm.dml | 2 +-
scripts/algorithms/stratstats.dml | 6 +-
scripts/datagen/genRandData4ChisquaredTest.dml | 4 +-
scripts/datagen/genRandData4FTest.dml | 6 +-
.../datagen/genRandData4LinearReg_LTstats.dml | 10 +-
scripts/datagen/genRandData4LogReg_LTstats.dml | 10 +-
scripts/datagen/genRandData4NMF.dml | 8 +-
scripts/datagen/genRandData4NMFBlockwise.dml | 8 +-
.../apply-transform/apply-transform.dml | 38 +++----
.../apply-transform/apply-transform.pydml | 38 +++----
.../applications/arima_box-jenkins/arima.dml | 20 ++--
.../applications/arima_box-jenkins/arima.pydml | 20 ++--
.../applications/cspline/CsplineCG.pydml | 2 +-
.../applications/cspline/CsplineDS.pydml | 2 +-
.../scripts/applications/ctableStats/ctci.dml | 2 +-
.../applications/ctableStats/stratstats.dml | 6 +-
.../applications/ctableStats/wilson_score.dml | 38 +++----
.../applications/descriptivestats/OddsRatio.dml | 8 +-
src/test/scripts/applications/glm/GLM.dml | 6 +-
src/test/scripts/applications/glm/GLM.pydml | 6 +-
src/test/scripts/applications/id3/id3.dml | 24 ++---
src/test/scripts/applications/id3/id3.pydml | 24 ++---
.../applications/impute/imputeGaussMCMC.dml | 54 +++++-----
.../impute/imputeGaussMCMC.nogradient.dml | 52 +++++-----
.../applications/impute/old/imputeGaussMCMC.dml | 42 ++++----
src/test/scripts/applications/impute/tmp.dml | 8 +-
.../impute/wfundInputGenerator1.dml | 2 +-
.../impute/wfundInputGenerator2.dml | 24 ++---
.../applications/linearLogReg/LinearLogReg.dml | 34 +++----
.../linearLogReg/LinearLogReg.pydml | 34 +++----
.../linear_regression/LinearRegression.dml | 2 +-
src/test/scripts/applications/m-svm/m-svm.dml | 2 +-
src/test/scripts/applications/m-svm/m-svm.pydml | 2 +-
.../applications/mdabivar/MDABivariateStats.dml | 10 +-
.../mdabivar/MDABivariateStats.pydml | 10 +-
.../applications/parfor/parfor_bivariate0.dml | 8 +-
.../applications/parfor/parfor_bivariate1.dml | 8 +-
.../applications/parfor/parfor_bivariate2.dml | 8 +-
.../applications/parfor/parfor_bivariate3.dml | 8 +-
.../applications/parfor/parfor_bivariate4.dml | 8 +-
.../applications/parfor/parfor_univariate0.dml | 2 +-
.../applications/parfor/parfor_univariate1.dml | 2 +-
.../applications/parfor/parfor_univariate4.dml | 2 +-
.../validation/CV_LogisticRegression.dml | 100 +++++++++----------
.../validation/CV_MultiClassSVM.sasha.dml | 12 +--
.../validation/LinearLogisticRegression.dml | 34 +++----
.../genRandData4LogisticRegression.dml | 8 +-
src/test/scripts/functions/gdfo/LinregCG.dml | 2 +-
.../functions/jmlc/reuse-glm-predict.dml | 18 ++--
.../functions/misc/IPAUnknownRecursion.dml | 4 +-
.../scripts/functions/misc/dt_change_4b.dml | 2 +-
.../scripts/functions/misc/dt_change_4c.dml | 2 +-
.../scripts/functions/misc/dt_change_4f.dml | 2 +-
src/test/scripts/functions/parfor/parfor35.dml | 2 +-
src/test/scripts/functions/parfor/parfor48b.dml | 2 +-
src/test/scripts/functions/parfor/parfor6.dml | 2 +-
src/test/scripts/functions/parfor/parfor7.dml | 2 +-
src/test/scripts/functions/parfor/parfor8.dml | 2 +-
src/test/scripts/functions/parfor/parfor9.dml | 2 +-
.../functions/parfor/parfor_optimizer2.dml | 12 +--
.../parfor/parfor_threadid_recompile1.dml | 8 +-
.../parfor/parfor_threadid_recompile2.dml | 8 +-
.../functions/recompile/for_recompile.dml | 2 +-
.../functions/recompile/if_recompile.dml | 2 +-
.../functions/recompile/parfor_recompile.dml | 2 +-
.../functions/recompile/while_recompile.dml | 2 +-
.../functions/unary/matrix/CastAsScalarTest.dml | 2 +-
.../scripts/functions/unary/matrix/eigen.dml | 4 +-
src/test/scripts/functions/unary/matrix/qr.dml | 2 +-
77 files changed, 475 insertions(+), 475 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/GLM-predict.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/GLM-predict.dml b/scripts/algorithms/GLM-predict.dml
index 5e998e3..355928b 100644
--- a/scripts/algorithms/GLM-predict.dml
+++ b/scripts/algorithms/GLM-predict.dml
@@ -312,15 +312,15 @@ if (fileY != " ")
str = append (str, "DEVIANCE_G2_PVAL,,TRUE," + G2_scaled_pValue);
for (i in 1:ncol(Y)) {
- str = append (str, "AVG_TOT_Y," + i + ",," + castAsScalar (avg_tot_Y [1, i]));
- str = append (str, "STDEV_TOT_Y," + i + ",," + castAsScalar (sqrt (var_tot_Y [1, i])));
- str = append (str, "AVG_RES_Y," + i + ",," + castAsScalar (avg_res_Y [1, i]));
- str = append (str, "STDEV_RES_Y," + i + ",," + castAsScalar (sqrt (var_res_Y [1, i])));
- str = append (str, "PRED_STDEV_RES," + i + ",TRUE," + castAsScalar (sqrt (predicted_avg_var_res_Y [1, i])));
- str = append (str, "PLAIN_R2," + i + ",," + castAsScalar (plain_R2 [1, i]));
- str = append (str, "ADJUSTED_R2," + i + ",," + castAsScalar (adjust_R2 [1, i]));
- str = append (str, "PLAIN_R2_NOBIAS," + i + ",," + castAsScalar (plain_R2_nobias [1, i]));
- str = append (str, "ADJUSTED_R2_NOBIAS," + i + ",," + castAsScalar (adjust_R2_nobias [1, i]));
+ str = append (str, "AVG_TOT_Y," + i + ",," + as.scalar (avg_tot_Y [1, i]));
+ str = append (str, "STDEV_TOT_Y," + i + ",," + as.scalar (sqrt (var_tot_Y [1, i])));
+ str = append (str, "AVG_RES_Y," + i + ",," + as.scalar (avg_res_Y [1, i]));
+ str = append (str, "STDEV_RES_Y," + i + ",," + as.scalar (sqrt (var_res_Y [1, i])));
+ str = append (str, "PRED_STDEV_RES," + i + ",TRUE," + as.scalar (sqrt (predicted_avg_var_res_Y [1, i])));
+ str = append (str, "PLAIN_R2," + i + ",," + as.scalar (plain_R2 [1, i]));
+ str = append (str, "ADJUSTED_R2," + i + ",," + as.scalar (adjust_R2 [1, i]));
+ str = append (str, "PLAIN_R2_NOBIAS," + i + ",," + as.scalar (plain_R2_nobias [1, i]));
+ str = append (str, "ADJUSTED_R2_NOBIAS," + i + ",," + as.scalar (adjust_R2_nobias [1, i]));
}
if (fileO != " ") {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/GLM.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/GLM.dml b/scripts/algorithms/GLM.dml
index 32a55f8..16008e6 100644
--- a/scripts/algorithms/GLM.dml
+++ b/scripts/algorithms/GLM.dml
@@ -453,7 +453,7 @@ if (intercept_status == 2) {
write (beta_out, fileB, format=fmtB);
if (intercept_status == 1 | intercept_status == 2) {
- intercept_value = castAsScalar (beta_out [num_features, 1]);
+ intercept_value = as.scalar (beta_out [num_features, 1]);
beta_noicept = beta_out [1 : (num_features - 1), 1];
} else {
beta_noicept = beta_out [1 : num_features, 1];
@@ -461,9 +461,9 @@ if (intercept_status == 1 | intercept_status == 2) {
min_beta = min (beta_noicept);
max_beta = max (beta_noicept);
tmp_i_min_beta = rowIndexMin (t(beta_noicept))
-i_min_beta = castAsScalar (tmp_i_min_beta [1, 1]);
+i_min_beta = as.scalar (tmp_i_min_beta [1, 1]);
tmp_i_max_beta = rowIndexMax (t(beta_noicept))
-i_max_beta = castAsScalar (tmp_i_max_beta [1, 1]);
+i_max_beta = as.scalar (tmp_i_max_beta [1, 1]);
##### OVER-DISPERSION PART #####
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/Kmeans-predict.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/Kmeans-predict.dml b/scripts/algorithms/Kmeans-predict.dml
index 5bd78bd..3823234 100644
--- a/scripts/algorithms/Kmeans-predict.dml
+++ b/scripts/algorithms/Kmeans-predict.dml
@@ -266,19 +266,19 @@ if (num_records != nrow (prY) | ncol (spY) != 1 | ncol (prY) != 1) {
for (i in 1 : nrow (spY_cids))
{
- cid = as.integer (castAsScalar (spY_cids [i, 1]));
- pct = castAsScalar (rounded_percentages [i, 1]);
+ cid = as.integer (as.scalar (spY_cids [i, 1]));
+ pct = as.scalar (rounded_percentages [i, 1]);
space_pct = ""; if (pct < 100) {space_pct = " ";} if (pct < 10) {space_pct = " ";}
print ("Category " + cid +
- ": best pred. cluster is " + as.integer (castAsScalar (prY_cids [i, 1])) +
- "; full count = " + as.integer (castAsScalar (full_counts [i, 1])) +
+ ": best pred. cluster is " + as.integer (as.scalar (prY_cids [i, 1])) +
+ "; full count = " + as.integer (as.scalar (full_counts [i, 1])) +
", matching count = " + space_pct + pct + "% (" +
- as.integer (castAsScalar (matching_counts [i, 1])) + ")");
+ as.integer (as.scalar (matching_counts [i, 1])) + ")");
- str = append (str, "SPEC_TO_PRED," + cid + "," + castAsScalar (prY_cids [i, 1]));
- str = append (str, "SPEC_FULL_CT," + cid + "," + castAsScalar (full_counts [i, 1]));
- str = append (str, "SPEC_MATCH_CT," + cid + "," + castAsScalar (matching_counts [i, 1]));
- str = append (str, "SPEC_MATCH_PC," + cid + "," + castAsScalar (rounded_percentages [i, 1]));
+ str = append (str, "SPEC_TO_PRED," + cid + "," + as.scalar (prY_cids [i, 1]));
+ str = append (str, "SPEC_FULL_CT," + cid + "," + as.scalar (full_counts [i, 1]));
+ str = append (str, "SPEC_MATCH_CT," + cid + "," + as.scalar (matching_counts [i, 1]));
+ str = append (str, "SPEC_MATCH_PC," + cid + "," + as.scalar (rounded_percentages [i, 1]));
}
[prY_cids, spY_cids, full_counts, matching_counts, rounded_percentages] =
@@ -292,19 +292,19 @@ if (num_records != nrow (prY) | ncol (spY) != 1 | ncol (prY) != 1) {
for (i in 1 : nrow (prY_cids))
{
- cid = as.integer (castAsScalar (prY_cids [i, 1]));
- pct = castAsScalar (rounded_percentages [i, 1]);
+ cid = as.integer (as.scalar (prY_cids [i, 1]));
+ pct = as.scalar (rounded_percentages [i, 1]);
space_pct = ""; if (pct < 100) {space_pct = " ";} if (pct < 10) {space_pct = " ";}
print ("Cluster " + cid +
- ": best spec. categ. is " + as.integer (castAsScalar (spY_cids [i, 1])) +
- "; full count = " + as.integer (castAsScalar (full_counts [i, 1])) +
+ ": best spec. categ. is " + as.integer (as.scalar (spY_cids [i, 1])) +
+ "; full count = " + as.integer (as.scalar (full_counts [i, 1])) +
", matching count = " + space_pct + pct + "% (" +
- as.integer (castAsScalar (matching_counts [i, 1])) + ")");
+ as.integer (as.scalar (matching_counts [i, 1])) + ")");
- str = append (str, "PRED_TO_SPEC," + cid + "," + castAsScalar (spY_cids [i, 1]));
- str = append (str, "PRED_FULL_CT," + cid + "," + castAsScalar (full_counts [i, 1]));
- str = append (str, "PRED_MATCH_CT," + cid + "," + castAsScalar (matching_counts [i, 1]));
- str = append (str, "PRED_MATCH_PC," + cid + "," + castAsScalar (rounded_percentages [i, 1]));
+ str = append (str, "PRED_TO_SPEC," + cid + "," + as.scalar (spY_cids [i, 1]));
+ str = append (str, "PRED_FULL_CT," + cid + "," + as.scalar (full_counts [i, 1]));
+ str = append (str, "PRED_MATCH_CT," + cid + "," + as.scalar (matching_counts [i, 1]));
+ str = append (str, "PRED_MATCH_PC," + cid + "," + as.scalar (rounded_percentages [i, 1]));
}
print (" ");
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/Kmeans.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/Kmeans.dml b/scripts/algorithms/Kmeans.dml
index 2887baa..8717473 100644
--- a/scripts/algorithms/Kmeans.dml
+++ b/scripts/algorithms/Kmeans.dml
@@ -190,11 +190,11 @@ termination_bitmap = matrix (0, rows = num_runs, cols = 3);
termination_bitmap_raw = table (seq (1, num_runs, 1), termination_code);
termination_bitmap [, 1 : ncol(termination_bitmap_raw)] = termination_bitmap_raw;
termination_stats = colSums (termination_bitmap);
-print ("Number of successful runs = " + as.integer (castAsScalar (termination_stats [1, 1])));
-print ("Number of incomplete runs = " + as.integer (castAsScalar (termination_stats [1, 2])));
-print ("Number of failed runs (with lost centroids) = " + as.integer (castAsScalar (termination_stats [1, 3])));
+print ("Number of successful runs = " + as.integer (as.scalar (termination_stats [1, 1])));
+print ("Number of incomplete runs = " + as.integer (as.scalar (termination_stats [1, 2])));
+print ("Number of failed runs (with lost centroids) = " + as.integer (as.scalar (termination_stats [1, 3])));
-num_successful_runs = castAsScalar (termination_stats [1, 1]);
+num_successful_runs = as.scalar (termination_stats [1, 1]);
if (num_successful_runs > 0) {
final_wcss_successful = final_wcss * termination_bitmap [, 1];
worst_wcss = max (final_wcss_successful);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/StepGLM.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/StepGLM.dml b/scripts/algorithms/StepGLM.dml
index 10737ff..a8c8820 100644
--- a/scripts/algorithms/StepGLM.dml
+++ b/scripts/algorithms/StepGLM.dml
@@ -498,7 +498,7 @@ glm = function (Matrix[Double] X, Matrix[Double] Y, Int intercept_status, Double
if (intercept_status == 1 | intercept_status == 2) {
- intercept_value = castAsScalar (beta_out [num_features, 1]);
+ intercept_value = as.scalar (beta_out [num_features, 1]);
beta_noicept = beta_out [1 : (num_features - 1), 1];
} else {
beta_noicept = beta_out [1 : num_features, 1];
@@ -506,9 +506,9 @@ glm = function (Matrix[Double] X, Matrix[Double] Y, Int intercept_status, Double
min_beta = min (beta_noicept);
max_beta = max (beta_noicept);
tmp_i_min_beta = rowIndexMin (t(beta_noicept))
- i_min_beta = castAsScalar (tmp_i_min_beta [1, 1]);
+ i_min_beta = as.scalar (tmp_i_min_beta [1, 1]);
tmp_i_max_beta = rowIndexMax (t(beta_noicept))
- i_max_beta = castAsScalar (tmp_i_max_beta [1, 1]);
+ i_max_beta = as.scalar (tmp_i_max_beta [1, 1]);
##### OVER-DISPERSION PART #####
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/Univar-Stats.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/Univar-Stats.dml b/scripts/algorithms/Univar-Stats.dml
index 404e002..525118c 100644
--- a/scripts/algorithms/Univar-Stats.dml
+++ b/scripts/algorithms/Univar-Stats.dml
@@ -69,7 +69,7 @@ parfor(i in 1:n, check=0) {
# project out the i^th column
F = A[,i];
- kind = castAsScalar(K[1,i]);
+ kind = as.scalar(K[1,i]);
if ( kind == 1 ) {
#print("[" + i + "] Scale");
@@ -149,7 +149,7 @@ parfor(i in 1:n, check=0) {
if (consoleOutput == TRUE) {
for(i in 1:n) {
print("-------------------------------------------------");
- kind = castAsScalar(K[1,i]);
+ kind = as.scalar(K[1,i]);
if (kind == 1) {
print("Feature [" + i + "]: Scale");
print(" (01) Minimum | " + as.scalar(baseStats[1,i]));
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/bivar-stats.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/bivar-stats.dml b/scripts/algorithms/bivar-stats.dml
index 99549dc..8f7b6c1 100644
--- a/scripts/algorithms/bivar-stats.dml
+++ b/scripts/algorithms/bivar-stats.dml
@@ -67,13 +67,13 @@ num_nominal_scale_tests = 0
pair2row = matrix(0, rows=numPairs, cols=2)
for( i in 1:s1size, check=0) {
- pre_a1 = castAsScalar(S1[1,i]);
- pre_k1 = castAsScalar(K1[1,i]);
+ pre_a1 = as.scalar(S1[1,i]);
+ pre_k1 = as.scalar(K1[1,i]);
for( j in 1:s2size, check=0) {
pre_pairID = (i-1)*s2size+j;
- pre_a2 = castAsScalar(S2[1,j]);
- pre_k2 = castAsScalar(K2[1,j]);
+ pre_a2 = as.scalar(S2[1,j]);
+ pre_k2 = as.scalar(K2[1,j]);
if (pre_k1 == pre_k2) {
if (pre_k1 == 1) {
@@ -167,18 +167,18 @@ maxDomain = as.integer(maxDomainSize);
if(error_flag) stop(debug_str);
parfor( i in 1:s1size, check=0) {
- a1 = castAsScalar(S1[1,i]);
- k1 = castAsScalar(K1[1,i]);
+ a1 = as.scalar(S1[1,i]);
+ k1 = as.scalar(K1[1,i]);
A1 = D[,a1];
parfor( j in 1:s2size, check=0) {
pairID = (i-1)*s2size+j;
- a2 = castAsScalar(S2[1,j]);
- k2 = castAsScalar(K2[1,j]);
+ a2 = as.scalar(S2[1,j]);
+ k2 = as.scalar(K2[1,j]);
A2 = D[,a2];
- rowid1 = castAsScalar(pair2row[pairID, 1])
- rowid2 = castAsScalar(pair2row[pairID, 2])
+ rowid1 = as.scalar(pair2row[pairID, 1])
+ rowid2 = as.scalar(pair2row[pairID, 2])
if (k1 == k2) {
if (k1 == 1) {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/l2-svm-predict.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/l2-svm-predict.dml b/scripts/algorithms/l2-svm-predict.dml
index a4d6fff..cace79f 100644
--- a/scripts/algorithms/l2-svm-predict.dml
+++ b/scripts/algorithms/l2-svm-predict.dml
@@ -54,7 +54,7 @@ w = w[1:(nrow(w)-4),]
b = 0.0
if(intercept == 1)
- b = castAsScalar(w[nrow(w),1])
+ b = as.scalar(w[nrow(w),1])
scores = b + (X %*% w[1:ncol(X),])
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/m-svm.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/m-svm.dml b/scripts/algorithms/m-svm.dml
index 4142ac1..4224d26 100644
--- a/scripts/algorithms/m-svm.dml
+++ b/scripts/algorithms/m-svm.dml
@@ -170,7 +170,7 @@ write(w, $model, format=cmdLine_fmt)
debug_str = "# Class, Iter, Obj"
for(iter_class in 1:ncol(debug_mat)){
for(iter in 1:nrow(debug_mat)){
- obj = castAsScalar(debug_mat[iter, iter_class])
+ obj = as.scalar(debug_mat[iter, iter_class])
if(obj != -1)
debug_str = append(debug_str, iter_class + "," + iter + "," + obj)
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/algorithms/stratstats.dml
----------------------------------------------------------------------
diff --git a/scripts/algorithms/stratstats.dml b/scripts/algorithms/stratstats.dml
index 2b7425d..d380220 100644
--- a/scripts/algorithms/stratstats.dml
+++ b/scripts/algorithms/stratstats.dml
@@ -375,9 +375,9 @@ fStat_tailprob = function (Matrix[double] fStat, Matrix[double] df_1, Matrix[dou
tailprob = fStat;
for (i in 1:nrow(fStat)) {
for (j in 1:ncol(fStat)) {
- q = castAsScalar (fStat [i, j]);
- d1 = castAsScalar (df_1 [i, j]);
- d2 = castAsScalar (df_2 [i, j]);
+ q = as.scalar (fStat [i, j]);
+ d1 = as.scalar (df_1 [i, j]);
+ d2 = as.scalar (df_2 [i, j]);
if (d1 >= 1 & d2 >= 1 & q >= 0.0) {
tailprob [i, j] = pf(target = q, df1 = d1, df2 = d2, lower.tail=FALSE);
} else {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4ChisquaredTest.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4ChisquaredTest.dml b/scripts/datagen/genRandData4ChisquaredTest.dml
index e25adf2..4709843 100644
--- a/scripts/datagen/genRandData4ChisquaredTest.dml
+++ b/scripts/datagen/genRandData4ChisquaredTest.dml
@@ -65,14 +65,14 @@ one = Rand(rows=1, cols=1, min=1.0, max=1.0, pdf="uniform", seed=0)
data = Rand(rows=numSamples, cols=2, min=0.0, max=0.0, pdf="uniform", seed=0)
parfor(s in 1:numSamples){
r_mat = Rand(rows=1, cols=1, min=0.0, max=1.0, pdf="uniform", seed=0)
- r = castAsScalar(r_mat)
+ r = as.scalar(r_mat)
cat1 = -1
cat2 = -1
continue = 1
for(i in 1:numCategories1){
for(j in 1:numCategories2){
- cdf = castAsScalar(oCDF[i,j])
+ cdf = as.scalar(oCDF[i,j])
if(continue == 1 & r <= cdf){
cat1 = i
cat2 = j
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4FTest.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4FTest.dml b/scripts/datagen/genRandData4FTest.dml
index bdd33b9..9f0e1d6 100644
--- a/scripts/datagen/genRandData4FTest.dml
+++ b/scripts/datagen/genRandData4FTest.dml
@@ -50,7 +50,7 @@ one = Rand(rows=1, cols=1, min=1.0, max=1.0, pdf="uniform")
copy_start_index = numActualGroups+1
parfor(i in copy_start_index:numGroups){
r = Rand(rows=1, cols=1, min=1.0, max=numActualGroups, pdf="uniform", seed=0)
- j = castAsScalar(round(r))
+ j = as.scalar(round(r))
permut[j,i] = one
}
@@ -77,12 +77,12 @@ for(i in 2:numGroups){
data = Rand(rows=numSamples, cols=1, min=0.0, max=0.0, pdf="uniform")
parfor(i in 1:numSamples){
r_mat = Rand(rows=1, cols=1, min=0.0, max=1.0, pdf="uniform", seed=0)
- r1 = castAsScalar(r_mat)
+ r1 = as.scalar(r_mat)
g = -1
continue = 1
for(k in 1:numGroups){
- cdf = castAsScalar(cntCDFs[k,1])
+ cdf = as.scalar(cntCDFs[k,1])
if(continue==1 & r1<=cdf){
g = k
continue=0
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4LinearReg_LTstats.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4LinearReg_LTstats.dml b/scripts/datagen/genRandData4LinearReg_LTstats.dml
index 0bc187a..e4e8384 100644
--- a/scripts/datagen/genRandData4LinearReg_LTstats.dml
+++ b/scripts/datagen/genRandData4LinearReg_LTstats.dml
@@ -92,13 +92,13 @@ actual_meanLT = colSums (LT) / numSamples;
actual_sigmaLT = sqrt (colSums ((LT - ones %*% actual_meanLT)^2) / numSamples);
for (i in 1:(numCategories - 1)) {
- if (castAsScalar (new_sigmaLT [1, i]) == castAsScalar (sigmaLT [1, i])) {
- print ("Category " + i + ": Intercept = " + castAsScalar (b_intercept [1, i]));
+ if (as.scalar (new_sigmaLT [1, i]) == as.scalar (sigmaLT [1, i])) {
+ print ("Category " + i + ": Intercept = " + as.scalar (b_intercept [1, i]));
} else {
- print ("Category " + i + ": Intercept = " + castAsScalar (b_intercept [1, i]) + ", st.dev.(LT) relaxed from " + castAsScalar (sigmaLT [1, i]));
+ print ("Category " + i + ": Intercept = " + as.scalar (b_intercept [1, i]) + ", st.dev.(LT) relaxed from " + as.scalar (sigmaLT [1, i]));
}
- print (" Wanted LT mean = " + castAsScalar (meanLT [1, i]) + ", st.dev. = " + castAsScalar (new_sigmaLT [1, i]));
- print (" Actual LT mean = " + castAsScalar (actual_meanLT [1, i]) + ", st.dev. = " + castAsScalar (actual_sigmaLT [1, i]));
+ print (" Wanted LT mean = " + as.scalar (meanLT [1, i]) + ", st.dev. = " + as.scalar (new_sigmaLT [1, i]));
+ print (" Actual LT mean = " + as.scalar (actual_meanLT [1, i]) + ", st.dev. = " + as.scalar (actual_sigmaLT [1, i]));
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4LogReg_LTstats.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4LogReg_LTstats.dml b/scripts/datagen/genRandData4LogReg_LTstats.dml
index 2ec5aef..1797f4f 100644
--- a/scripts/datagen/genRandData4LogReg_LTstats.dml
+++ b/scripts/datagen/genRandData4LogReg_LTstats.dml
@@ -91,13 +91,13 @@ actual_meanLT = colSums (LT) / numSamples;
actual_sigmaLT = sqrt (colSums ((LT - ones %*% actual_meanLT)^2) / numSamples);
for (i in 1:(numCategories - 1)) {
- if (castAsScalar (new_sigmaLT [1, i]) == castAsScalar (sigmaLT [1, i])) {
- print ("Category " + i + ": Intercept = " + castAsScalar (b_intercept [1, i]));
+ if (as.scalar (new_sigmaLT [1, i]) == as.scalar (sigmaLT [1, i])) {
+ print ("Category " + i + ": Intercept = " + as.scalar (b_intercept [1, i]));
} else {
- print ("Category " + i + ": Intercept = " + castAsScalar (b_intercept [1, i]) + ", st.dev.(LT) relaxed from " + castAsScalar (sigmaLT [1, i]));
+ print ("Category " + i + ": Intercept = " + as.scalar (b_intercept [1, i]) + ", st.dev.(LT) relaxed from " + as.scalar (sigmaLT [1, i]));
}
- print (" Wanted LT mean = " + castAsScalar (meanLT [1, i]) + ", st.dev. = " + castAsScalar (new_sigmaLT [1, i]));
- print (" Actual LT mean = " + castAsScalar (actual_meanLT [1, i]) + ", st.dev. = " + castAsScalar (actual_sigmaLT [1, i]));
+ print (" Wanted LT mean = " + as.scalar (meanLT [1, i]) + ", st.dev. = " + as.scalar (new_sigmaLT [1, i]));
+ print (" Actual LT mean = " + as.scalar (actual_meanLT [1, i]) + ", st.dev. = " + as.scalar (actual_sigmaLT [1, i]));
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4NMF.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4NMF.dml b/scripts/datagen/genRandData4NMF.dml
index cf18430..87e3f47 100644
--- a/scripts/datagen/genRandData4NMF.dml
+++ b/scripts/datagen/genRandData4NMF.dml
@@ -80,7 +80,7 @@ parfor(i in 1:numDocuments){
r_w = Rand(rows=numWordsPerDoc, cols=1, min=0, max=1, pdf="uniform", seed=0)
for(j in 1:numWordsPerDoc){
- rz = castAsScalar(r_z[j,1])
+ rz = as.scalar(r_z[j,1])
continue = 1
z = -1
@@ -88,7 +88,7 @@ parfor(i in 1:numDocuments){
#z=1
for(k1 in 1:numTopics){
- prob = castAsScalar(docTopic[1,k1])
+ prob = as.scalar(docTopic[1,k1])
if(continue==1 & rz <= prob){
z=k1
continue=0
@@ -100,7 +100,7 @@ parfor(i in 1:numDocuments){
z = numTopics
}
- rw = castAsScalar(r_w[j,1])
+ rw = as.scalar(r_w[j,1])
continue = 1
w = -1
@@ -108,7 +108,7 @@ parfor(i in 1:numDocuments){
#w = 1
for(k2 in 1:numFeatures){
- prob = castAsScalar(topicDistributions[z,k2])
+ prob = as.scalar(topicDistributions[z,k2])
if(continue == 1 & rw <= prob){
w = k2
continue = 0
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/scripts/datagen/genRandData4NMFBlockwise.dml
----------------------------------------------------------------------
diff --git a/scripts/datagen/genRandData4NMFBlockwise.dml b/scripts/datagen/genRandData4NMFBlockwise.dml
index e3fd67f..06b8057 100644
--- a/scripts/datagen/genRandData4NMFBlockwise.dml
+++ b/scripts/datagen/genRandData4NMFBlockwise.dml
@@ -88,7 +88,7 @@ for( k in seq(1,numDocuments,blocksize) )
r_w = Rand(rows=numWordsPerDoc, cols=1, min=0, max=1, pdf="uniform", seed=0)
for(j in 1:numWordsPerDoc){
- rz = castAsScalar(r_z[j,1])
+ rz = as.scalar(r_z[j,1])
continue = 1
z = -1
@@ -96,7 +96,7 @@ for( k in seq(1,numDocuments,blocksize) )
#z=1
for(k1 in 1:numTopics){
- prob = castAsScalar(docTopic[1,k1])
+ prob = as.scalar(docTopic[1,k1])
if(continue==1 & rz <= prob){
z=k1
continue=0
@@ -108,7 +108,7 @@ for( k in seq(1,numDocuments,blocksize) )
z = numTopics
}
- rw = castAsScalar(r_w[j,1])
+ rw = as.scalar(r_w[j,1])
continue = 1
w = -1
@@ -116,7 +116,7 @@ for( k in seq(1,numDocuments,blocksize) )
#w = 1
for(k2 in 1:numFeatures){
- prob = castAsScalar(topicDistributions[z,k2])
+ prob = as.scalar(topicDistributions[z,k2])
if(continue == 1 & rw <= prob){
w = k2
continue = 0
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/apply-transform/apply-transform.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/apply-transform/apply-transform.dml b/src/test/scripts/applications/apply-transform/apply-transform.dml
index fdd85c7..5110cb0 100644
--- a/src/test/scripts/applications/apply-transform/apply-transform.dml
+++ b/src/test/scripts/applications/apply-transform/apply-transform.dml
@@ -52,7 +52,7 @@ if(cmdLine_missing_value_maps != " "){
missing_indicator_mat = original_X[,(last_data_col+1):ncol(original_X)]
parfor(i in 1:nrow(missing_val_maps), check=0){
- attr_index_mv = castAsScalar(missing_val_maps[i,1])
+ attr_index_mv = as.scalar(missing_val_maps[i,1])
attrinfo[attr_index_mv,1] = i
attrinfo[attr_index_mv,2] = missing_val_maps[i,2]
}
@@ -61,7 +61,7 @@ if(cmdLine_missing_value_maps != " "){
if(cmdLine_bin_defns != " "){
bin_defns = read(cmdLine_bin_defns)
parfor(i in 1:nrow(bin_defns), check=0){
- attr_index_bin = castAsScalar(bin_defns[i,1])
+ attr_index_bin = as.scalar(bin_defns[i,1])
attrinfo[attr_index_bin,3] = bin_defns[i,4]
attrinfo[attr_index_bin,4] = bin_defns[i,2]
attrinfo[attr_index_bin,5] = bin_defns[i,3]
@@ -71,7 +71,7 @@ if(cmdLine_bin_defns != " "){
if(cmdLine_dummy_code_maps != " "){
dummy_code_maps = read(cmdLine_dummy_code_maps)
parfor(i in 1:nrow(dummy_code_maps), check=0){
- attr_index_dc = castAsScalar(dummy_code_maps[i,1])
+ attr_index_dc = as.scalar(dummy_code_maps[i,1])
attrinfo[attr_index_dc,6] = dummy_code_maps[i,2]
attrinfo[attr_index_dc,7] = dummy_code_maps[i,3]
}
@@ -83,31 +83,31 @@ if(cmdLine_dummy_code_maps != " "){
if(cmdLine_normalization_maps != " "){
normalization_map = read(cmdLine_normalization_maps)
parfor(i in 1:nrow(normalization_map), check=0){
- attr_index_normalization = castAsScalar(normalization_map[i,1])
+ attr_index_normalization = as.scalar(normalization_map[i,1])
attrinfo[attr_index_normalization,8] = 1
- attrinfo[attr_index_normalization,9] = castAsScalar(normalization_map[i,2])
- attrinfo[attr_index_normalization,10] = castAsScalar(normalization_map[i,3])
+ attrinfo[attr_index_normalization,9] = as.scalar(normalization_map[i,2])
+ attrinfo[attr_index_normalization,10] = as.scalar(normalization_map[i,3])
}
}
#write(attrinfo, "binning/attrinfo.mtx", format="csv")
-cols_in_transformed_X = castAsScalar(attrinfo[nrow(attrinfo),6])
+cols_in_transformed_X = as.scalar(attrinfo[nrow(attrinfo),6])
new_X = matrix(0, rows=nrow(X), cols=cols_in_transformed_X)
log = matrix(0, rows=ncol(X), cols=2)
parfor(i in 1:ncol(X), check=0){
col = X[,i]
- mv_col_id = castAsScalar(attrinfo[i,1])
- global_mean = castAsScalar(attrinfo[i,2])
- num_bins = castAsScalar(attrinfo[i,3])
- bin_width = castAsScalar(attrinfo[i,4])
- min_val = castAsScalar(attrinfo[i,5])
- dummy_coding_beg_col = castAsScalar(attrinfo[i,6])
- dummy_coding_end_col = castAsScalar(attrinfo[i,7])
- normalization_needed = castAsScalar(attrinfo[i,8])
- normalization_mean = castAsScalar(attrinfo[i,9])
- normalization_std = castAsScalar(attrinfo[i,10])
+ mv_col_id = as.scalar(attrinfo[i,1])
+ global_mean = as.scalar(attrinfo[i,2])
+ num_bins = as.scalar(attrinfo[i,3])
+ bin_width = as.scalar(attrinfo[i,4])
+ min_val = as.scalar(attrinfo[i,5])
+ dummy_coding_beg_col = as.scalar(attrinfo[i,6])
+ dummy_coding_end_col = as.scalar(attrinfo[i,7])
+ normalization_needed = as.scalar(attrinfo[i,8])
+ normalization_mean = as.scalar(attrinfo[i,9])
+ normalization_std = as.scalar(attrinfo[i,10])
if(mv_col_id > 0){
# fill-in with global mean
@@ -150,7 +150,7 @@ write(new_X, $transformed_X, format="text")
s = "Warning Messages"
for(i in 1:nrow(log)){
- if(castAsScalar(log[i,1]) == 1)
- s = append(s, "Unseen value in column " + i + " (" + castAsScalar(log[i,2]) + ")")
+ if(as.scalar(log[i,1]) == 1)
+ s = append(s, "Unseen value in column " + i + " (" + as.scalar(log[i,2]) + ")")
}
write(s, $Log)
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/apply-transform/apply-transform.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/apply-transform/apply-transform.pydml b/src/test/scripts/applications/apply-transform/apply-transform.pydml
index f6c40dd..cc95e85 100644
--- a/src/test/scripts/applications/apply-transform/apply-transform.pydml
+++ b/src/test/scripts/applications/apply-transform/apply-transform.pydml
@@ -52,14 +52,14 @@ if(cmdLine_missing_value_maps != " "):
missing_indicator_mat = original_X[,(last_data_col+1):ncol(original_X)]
parfor(i in 1:nrow(missing_val_maps), check=0):
- attr_index_mv = castAsScalar(missing_val_maps[i,1])
+ attr_index_mv = scalar(missing_val_maps[i,1])
attrinfo[attr_index_mv,1] = i
attrinfo[attr_index_mv,2] = missing_val_maps[i,2]
if(cmdLine_bin_defns != " "):
bin_defns = read(cmdLine_bin_defns)
parfor(i in 1:nrow(bin_defns), check=0):
- attr_index_bin = castAsScalar(bin_defns[i,1])
+ attr_index_bin = scalar(bin_defns[i,1])
attrinfo[attr_index_bin,3] = bin_defns[i,4]
attrinfo[attr_index_bin,4] = bin_defns[i,2]
attrinfo[attr_index_bin,5] = bin_defns[i,3]
@@ -67,7 +67,7 @@ if(cmdLine_bin_defns != " "):
if(cmdLine_dummy_code_maps != " "):
dummy_code_maps = read(cmdLine_dummy_code_maps)
parfor(i in 1:nrow(dummy_code_maps), check=0):
- attr_index_dc = castAsScalar(dummy_code_maps[i,1])
+ attr_index_dc = scalar(dummy_code_maps[i,1])
attrinfo[attr_index_dc,6] = dummy_code_maps[i,2]
attrinfo[attr_index_dc,7] = dummy_code_maps[i,3]
else:
@@ -77,29 +77,29 @@ else:
if(cmdLine_normalization_maps != " "):
normalization_map = read(cmdLine_normalization_maps)
parfor(i in 1:nrow(normalization_map), check=0):
- attr_index_normalization = castAsScalar(normalization_map[i,1])
+ attr_index_normalization = scalar(normalization_map[i,1])
attrinfo[attr_index_normalization,8] = 1
- attrinfo[attr_index_normalization,9] = castAsScalar(normalization_map[i,2])
- attrinfo[attr_index_normalization,10] = castAsScalar(normalization_map[i,3])
+ attrinfo[attr_index_normalization,9] = scalar(normalization_map[i,2])
+ attrinfo[attr_index_normalization,10] = scalar(normalization_map[i,3])
#write(attrinfo, "binning/attrinfo.mtx", format="csv")
-cols_in_transformed_X = castAsScalar(attrinfo[nrow(attrinfo),6])
+cols_in_transformed_X = scalar(attrinfo[nrow(attrinfo),6])
new_X = full(0, rows=nrow(X), cols=cols_in_transformed_X)
log = full(0, rows=ncol(X), cols=2)
parfor(i in 1:ncol(X), check=0):
col = X[,i]
- mv_col_id = castAsScalar(attrinfo[i,1])
- global_mean = castAsScalar(attrinfo[i,2])
- num_bins = castAsScalar(attrinfo[i,3])
- bin_width = castAsScalar(attrinfo[i,4])
- min_val = castAsScalar(attrinfo[i,5])
- dummy_coding_beg_col = castAsScalar(attrinfo[i,6])
- dummy_coding_end_col = castAsScalar(attrinfo[i,7])
- normalization_needed = castAsScalar(attrinfo[i,8])
- normalization_mean = castAsScalar(attrinfo[i,9])
- normalization_std = castAsScalar(attrinfo[i,10])
+ mv_col_id = scalar(attrinfo[i,1])
+ global_mean = scalar(attrinfo[i,2])
+ num_bins = scalar(attrinfo[i,3])
+ bin_width = scalar(attrinfo[i,4])
+ min_val = scalar(attrinfo[i,5])
+ dummy_coding_beg_col = scalar(attrinfo[i,6])
+ dummy_coding_end_col = scalar(attrinfo[i,7])
+ normalization_needed = scalar(attrinfo[i,8])
+ normalization_mean = scalar(attrinfo[i,9])
+ normalization_std = scalar(attrinfo[i,10])
if(mv_col_id > 0):
# fill-in with global mean
@@ -140,7 +140,7 @@ save(new_X, $transformed_X, format="text")
s = "Warning Messages"
for(i in 1:nrow(log)):
- if(castAsScalar(log[i,1]) == 1):
- s = append(s, "Unseen value in column " + i + " (" + castAsScalar(log[i,2]) + ")")
+ if(scalar(log[i,1]) == 1):
+ s = append(s, "Unseen value in column " + i + " (" + scalar(log[i,2]) + ")")
save(s, $Log)
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/arima_box-jenkins/arima.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/arima_box-jenkins/arima.dml b/src/test/scripts/applications/arima_box-jenkins/arima.dml
index 73052e0..e21b75e 100644
--- a/src/test/scripts/applications/arima_box-jenkins/arima.dml
+++ b/src/test/scripts/applications/arima_box-jenkins/arima.dml
@@ -43,14 +43,14 @@ arima_css = function(Matrix[Double] w, Matrix[Double] X, Integer pIn, Integer P,
ma_ind_ns = P+pIn+i7
err_ind_ns = i7
ones_ns = Rand(rows=nrow(R)-err_ind_ns, cols=1, min=1, max=1)
- d_ns = ones_ns * castAsScalar(w[ma_ind_ns,1])
+ d_ns = ones_ns * as.scalar(w[ma_ind_ns,1])
R[1+err_ind_ns:nrow(R),1:ncol(R)-err_ind_ns] = R[1+err_ind_ns:nrow(R),1:ncol(R)-err_ind_ns] + diag(d_ns)
}
for(i8 in 1:Q){
ma_ind_s = P+pIn+qIn+i8
err_ind_s = s*i8
ones_s = Rand(rows=nrow(R)-err_ind_s, cols=1, min=1, max=1)
- d_s = ones_s * castAsScalar(w[ma_ind_s,1])
+ d_s = ones_s * as.scalar(w[ma_ind_s,1])
R[1+err_ind_s:nrow(R),1:ncol(R)-err_ind_s] = R[1+err_ind_s:nrow(R),1:ncol(R)-err_ind_s] + diag(d_s)
}
@@ -91,7 +91,7 @@ arima_css = function(Matrix[Double] w, Matrix[Double] X, Integer pIn, Integer P,
}
while(iter < max_iter & continue == 1){
q = Z%*%p
- alpha = norm_r2 / castAsScalar(t(p) %*% q)
+ alpha = norm_r2 / as.scalar(t(p) %*% q)
y_hat = y_hat + alpha * p
old_norm_r2 = norm_r2
r = r + alpha * q
@@ -203,26 +203,26 @@ parfor(i3 in 1:ncol(simplex)){
}
num_func_invoc = num_func_invoc + ncol(simplex)
-tol = 1.5 * 10^(-8) * castAsScalar(objvals[1,1])
+tol = 1.5 * 10^(-8) * as.scalar(objvals[1,1])
continue = 1
while(continue == 1 & num_func_invoc <= max_func_invoc) {
best_index = 1
worst_index = 1
for(i in 2:ncol(objvals)){
- this = castAsScalar(objvals[1,i])
- that = castAsScalar(objvals[1,best_index])
+ this = as.scalar(objvals[1,i])
+ that = as.scalar(objvals[1,best_index])
if(that > this){
best_index = i
}
- that = castAsScalar(objvals[1,worst_index])
+ that = as.scalar(objvals[1,worst_index])
if(that < this){
worst_index = i
}
}
- best_obj_val = castAsScalar(objvals[1,best_index])
- worst_obj_val = castAsScalar(objvals[1,worst_index])
+ best_obj_val = as.scalar(objvals[1,best_index])
+ worst_obj_val = as.scalar(objvals[1,worst_index])
if(worst_obj_val <= best_obj_val + tol){
continue = 0
}
@@ -257,7 +257,7 @@ while(continue == 1 & num_func_invoc <= max_func_invoc) {
obj_x_c_in = arima_css(x_c_in, Z, p, P, q, Q, s, useJacobi)
num_func_invoc = num_func_invoc + 1
- if(obj_x_c_in < castAsScalar(objvals[1,worst_index])){
+ if(obj_x_c_in < as.scalar(objvals[1,worst_index])){
simplex[,worst_index] = x_c_in
objvals[1,worst_index] = obj_x_c_in
}else{
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/arima_box-jenkins/arima.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/arima_box-jenkins/arima.pydml b/src/test/scripts/applications/arima_box-jenkins/arima.pydml
index 9b3387c..1df70d1 100644
--- a/src/test/scripts/applications/arima_box-jenkins/arima.pydml
+++ b/src/test/scripts/applications/arima_box-jenkins/arima.pydml
@@ -43,13 +43,13 @@ def arima_css(w:matrix[float], X:matrix[float], pIn: int, P: int, qIn: int, Q:in
ma_ind_ns = P+pIn+i7
err_ind_ns = i7
ones_ns = Rand(rows=nrow(R)-err_ind_ns, cols=1, min=1, max=1)
- d_ns = ones_ns * castAsScalar(w[ma_ind_ns,1])
+ d_ns = ones_ns * scalar(w[ma_ind_ns,1])
R[1+err_ind_ns:nrow(R),1:ncol(R)-err_ind_ns] = R[1+err_ind_ns:nrow(R),1:ncol(R)-err_ind_ns] + diag(d_ns)
for(i8 in 1:Q):
ma_ind_s = P+pIn+qIn+i8
err_ind_s = s*i8
ones_s = Rand(rows=nrow(R)-err_ind_s, cols=1, min=1, max=1)
- d_s = ones_s * castAsScalar(w[ma_ind_s,1])
+ d_s = ones_s * scalar(w[ma_ind_s,1])
R[1+err_ind_s:nrow(R),1:ncol(R)-err_ind_s] = R[1+err_ind_s:nrow(R),1:ncol(R)-err_ind_s] + diag(d_s)
#checking for strict diagonal dominance
@@ -86,7 +86,7 @@ def arima_css(w:matrix[float], X:matrix[float], pIn: int, P: int, qIn: int, Q:in
while(iter < max_iter & continue == 1):
q = dot(Z, p)
transpose_p = transpose(p)
- alpha = norm_r2 / castAsScalar(dot(transpose_p, q))
+ alpha = norm_r2 / scalar(dot(transpose_p, q))
y_hat = y_hat + alpha * p
old_norm_r2 = norm_r2
r = r + alpha * q
@@ -187,23 +187,23 @@ parfor(i3 in 1:ncol(simplex)):
num_func_invoc = num_func_invoc + ncol(simplex)
-tol = 1.5 * (10**-8) * castAsScalar(objvals[1,1])
+tol = 1.5 * (10**-8) * scalar(objvals[1,1])
continue = 1
while(continue == 1 & num_func_invoc <= max_func_invoc):
best_index = 1
worst_index = 1
for(i in 2:ncol(objvals)):
- this = castAsScalar(objvals[1,i])
- that = castAsScalar(objvals[1,best_index])
+ this = scalar(objvals[1,i])
+ that = scalar(objvals[1,best_index])
if(that > this):
best_index = i
- that = castAsScalar(objvals[1,worst_index])
+ that = scalar(objvals[1,worst_index])
if(that < this):
worst_index = i
- best_obj_val = castAsScalar(objvals[1,best_index])
- worst_obj_val = castAsScalar(objvals[1,worst_index])
+ best_obj_val = scalar(objvals[1,best_index])
+ worst_obj_val = scalar(objvals[1,worst_index])
if(worst_obj_val <= best_obj_val + tol):
continue = 0
@@ -235,7 +235,7 @@ while(continue == 1 & num_func_invoc <= max_func_invoc):
obj_x_c_in = arima_css(x_c_in, Z, p, P, q, Q, s, useJacobi)
num_func_invoc = num_func_invoc + 1
- if(obj_x_c_in < castAsScalar(objvals[1,worst_index])):
+ if(obj_x_c_in < scalar(objvals[1,worst_index])):
simplex[,worst_index] = x_c_in
objvals[1,worst_index] = obj_x_c_in
else:
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/cspline/CsplineCG.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/cspline/CsplineCG.pydml b/src/test/scripts/applications/cspline/CsplineCG.pydml
index 1c8daad..29c55a1 100644
--- a/src/test/scripts/applications/cspline/CsplineCG.pydml
+++ b/src/test/scripts/applications/cspline/CsplineCG.pydml
@@ -155,7 +155,7 @@ def interpSpline(x: float, X: matrix[float], Y: matrix[float], K: matrix[float])
qm = (1-t)*Y[i-1,1] + t*Y[i,1] + t*(1-t)*(a*(1-t)+b*t)
- q = castAsScalar(qm)
+ q = scalar(qm)
#solve Ax = b
# for CG our formulation is
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/cspline/CsplineDS.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/cspline/CsplineDS.pydml b/src/test/scripts/applications/cspline/CsplineDS.pydml
index 2b865b3..40847a7 100644
--- a/src/test/scripts/applications/cspline/CsplineDS.pydml
+++ b/src/test/scripts/applications/cspline/CsplineDS.pydml
@@ -134,7 +134,7 @@ def interpSpline(x: float, X: matrix[float], Y: matrix[float], K: matrix[float])
qm = (1-t)*Y[i-1,1] + t*Y[i,1] + t*(1-t)*(a*(1-t)+b*t)
- q = castAsScalar(qm)
+ q = scalar(qm)
#
# trunc the matrix by the specified amount in the specified direction.
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/ctableStats/ctci.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/ctableStats/ctci.dml b/src/test/scripts/applications/ctableStats/ctci.dml
index 4aa3ae3..f545138 100644
--- a/src/test/scripts/applications/ctableStats/ctci.dml
+++ b/src/test/scripts/applications/ctableStats/ctci.dml
@@ -112,7 +112,7 @@ for (iLabel in 1:numLabels)
print (" (partition & label) / (all label) ratios...");
- cntThisLabel = zeros + castAsScalar (cntLabels [iLabel, 1]);
+ cntThisLabel = zeros + as.scalar (cntLabels [iLabel, 1]);
[ratio2, left_conf_wilson2, right_conf_wilson2] =
wilson_confidence (cntThisLabel, cntPartitionsWithLabel);
[left_conf_exact2] = binomQuantile (cntThisLabel, cntPartitionsWithLabel_minus_1, big_alpha);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/ctableStats/stratstats.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/ctableStats/stratstats.dml b/src/test/scripts/applications/ctableStats/stratstats.dml
index 7eb1858..5d190e7 100644
--- a/src/test/scripts/applications/ctableStats/stratstats.dml
+++ b/src/test/scripts/applications/ctableStats/stratstats.dml
@@ -317,9 +317,9 @@ fStat_tailprob = function (Matrix[double] fStat, Matrix[double] df_1, Matrix[dou
tailprob = fStat;
for (i in 1:nrow(fStat)) {
for (j in 1:ncol(fStat)) {
- q = castAsScalar (fStat [i, j]);
- d1 = castAsScalar (df_1 [i, j]);
- d2 = castAsScalar (df_2 [i, j]);
+ q = as.scalar (fStat [i, j]);
+ d1 = as.scalar (df_1 [i, j]);
+ d2 = as.scalar (df_2 [i, j]);
if (d1 >= 1 & d2 >= 1 & q >= 0.0) {
tailprob [i, j] = pf (target = q, df1 = d1, df2 = d2, lower.tail=FALSE);
} else {
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/ctableStats/wilson_score.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/ctableStats/wilson_score.dml b/src/test/scripts/applications/ctableStats/wilson_score.dml
index 27d0899..51e31b1 100644
--- a/src/test/scripts/applications/ctableStats/wilson_score.dml
+++ b/src/test/scripts/applications/ctableStats/wilson_score.dml
@@ -37,7 +37,7 @@ source ("Binomial.dml");
# test_m = Rand (rows = 1, cols = 1, min = 0, max = 0);
# test_p = Rand (rows = 1, cols = 1, min = 0.00421, max = 0.00421);
# [alpha] = binomProb (test_n, test_m, test_p);
-# print ("TEST: Prob [Binom (" + castAsScalar (test_n) + ", " + castAsScalar (test_p) + ") <= " + castAsScalar (test_m) + "] = " + castAsScalar (alpha));
+# print ("TEST: Prob [Binom (" + as.scalar (test_n) + ", " + as.scalar (test_p) + ") <= " + as.scalar (test_m) + "] = " + as.scalar (alpha));
print ("BEGIN WILSON SCORE SCRIPT");
print ("Reading X...");
@@ -87,39 +87,39 @@ result [, 15] = r_m_n_exact;
print ("M / sum(M) RESULTS: Wilson, Exact");
for (i in 1:num_rows) {
- p1 = castAsScalar (round (result [i, 1] * 100000) / 1000);
- lw1 = castAsScalar (round (result [i, 2] * 100000) / 1000);
- rw1 = castAsScalar (round (result [i, 3] * 100000) / 1000);
- le1 = castAsScalar (round (result [i, 4] * 100000) / 1000);
- re1 = castAsScalar (round (result [i, 5] * 100000) / 1000);
+ p1 = as.scalar (round (result [i, 1] * 100000) / 1000);
+ lw1 = as.scalar (round (result [i, 2] * 100000) / 1000);
+ rw1 = as.scalar (round (result [i, 3] * 100000) / 1000);
+ le1 = as.scalar (round (result [i, 4] * 100000) / 1000);
+ re1 = as.scalar (round (result [i, 5] * 100000) / 1000);
print ("Row " + i + ": "
- + castAsScalar (M [i, 1]) + "/" + castAsScalar (sum_M [i, 1]) + " = "
+ + as.scalar (M [i, 1]) + "/" + as.scalar (sum_M [i, 1]) + " = "
+ p1 + "% [" + lw1 + "%, " + rw1 + "%] [" + le1 + "%, " + re1 + "%]");
}
print ("N / sum(N) RESULTS: Wilson, Exact");
for (i in 1:num_rows) {
- p2 = castAsScalar (round (result [i, 6] * 100000) / 1000);
- lw2 = castAsScalar (round (result [i, 7] * 100000) / 1000);
- rw2 = castAsScalar (round (result [i, 8] * 100000) / 1000);
- le2 = castAsScalar (round (result [i, 9] * 100000) / 1000);
- re2 = castAsScalar (round (result [i, 10] * 100000) / 1000);
+ p2 = as.scalar (round (result [i, 6] * 100000) / 1000);
+ lw2 = as.scalar (round (result [i, 7] * 100000) / 1000);
+ rw2 = as.scalar (round (result [i, 8] * 100000) / 1000);
+ le2 = as.scalar (round (result [i, 9] * 100000) / 1000);
+ re2 = as.scalar (round (result [i, 10] * 100000) / 1000);
print ("Row " + i + ": "
- + castAsScalar (N [i, 1]) + "/" + castAsScalar (sum_N [i, 1]) + " = "
+ + as.scalar (N [i, 1]) + "/" + as.scalar (sum_N [i, 1]) + " = "
+ p2 + "% [" + lw2 + "%, " + rw2 + "%] [" + le2 + "%, " + re2 + "%] ");
}
print ("M / N RESULTS: Wilson, Exact");
for (i in 1:num_rows) {
- p3 = castAsScalar (round (result [i, 11] * 100000) / 1000);
- lw3 = castAsScalar (round (result [i, 12] * 100000) / 1000);
- rw3 = castAsScalar (round (result [i, 13] * 100000) / 1000);
- le3 = castAsScalar (round (result [i, 14] * 100000) / 1000);
- re3 = castAsScalar (round (result [i, 15] * 100000) / 1000);
+ p3 = as.scalar (round (result [i, 11] * 100000) / 1000);
+ lw3 = as.scalar (round (result [i, 12] * 100000) / 1000);
+ rw3 = as.scalar (round (result [i, 13] * 100000) / 1000);
+ le3 = as.scalar (round (result [i, 14] * 100000) / 1000);
+ re3 = as.scalar (round (result [i, 15] * 100000) / 1000);
print ("Row " + i + ": "
- + castAsScalar (M [i, 1]) + "/" + castAsScalar ( N [i, 1]) + " = "
+ + as.scalar (M [i, 1]) + "/" + as.scalar ( N [i, 1]) + " = "
+ p3 + "% [" + lw3 + "%, " + rw3 + "%] [" + le3 + "%, " + re3 + "%] ");
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/descriptivestats/OddsRatio.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/descriptivestats/OddsRatio.dml b/src/test/scripts/applications/descriptivestats/OddsRatio.dml
index ae52e03..3d8c84b 100644
--- a/src/test/scripts/applications/descriptivestats/OddsRatio.dml
+++ b/src/test/scripts/applications/descriptivestats/OddsRatio.dml
@@ -61,10 +61,10 @@ else {
# Given a 2x2 contingency table, it computes oddsRatio and the corresponding confidence interval
pair_corr = function(Matrix[Double] A) return (Double oddsRatio, Double left_conf, Double right_conf, Double sd, Double chisquared, Double pvalue, Double crv, Double sigma_away, Double df) {
- a11 = castAsScalar(A[1,1]);
- a12 = castAsScalar(A[1,2]);
- a21 = castAsScalar(A[2,1]);
- a22 = castAsScalar(A[2,2]);
+ a11 = as.scalar(A[1,1]);
+ a12 = as.scalar(A[1,2]);
+ a21 = as.scalar(A[2,1]);
+ a22 = as.scalar(A[2,2]);
sd = sqrt(1/a11 + 1/a12 + 1/a21 + 1/a22);
oddsRatio = (a11*a22)/(a12*a21);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/glm/GLM.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/glm/GLM.dml b/src/test/scripts/applications/glm/GLM.dml
index 32a55f8..16008e6 100644
--- a/src/test/scripts/applications/glm/GLM.dml
+++ b/src/test/scripts/applications/glm/GLM.dml
@@ -453,7 +453,7 @@ if (intercept_status == 2) {
write (beta_out, fileB, format=fmtB);
if (intercept_status == 1 | intercept_status == 2) {
- intercept_value = castAsScalar (beta_out [num_features, 1]);
+ intercept_value = as.scalar (beta_out [num_features, 1]);
beta_noicept = beta_out [1 : (num_features - 1), 1];
} else {
beta_noicept = beta_out [1 : num_features, 1];
@@ -461,9 +461,9 @@ if (intercept_status == 1 | intercept_status == 2) {
min_beta = min (beta_noicept);
max_beta = max (beta_noicept);
tmp_i_min_beta = rowIndexMin (t(beta_noicept))
-i_min_beta = castAsScalar (tmp_i_min_beta [1, 1]);
+i_min_beta = as.scalar (tmp_i_min_beta [1, 1]);
tmp_i_max_beta = rowIndexMax (t(beta_noicept))
-i_max_beta = castAsScalar (tmp_i_max_beta [1, 1]);
+i_max_beta = as.scalar (tmp_i_max_beta [1, 1]);
##### OVER-DISPERSION PART #####
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/glm/GLM.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/glm/GLM.pydml b/src/test/scripts/applications/glm/GLM.pydml
index cb63302..e737a13 100644
--- a/src/test/scripts/applications/glm/GLM.pydml
+++ b/src/test/scripts/applications/glm/GLM.pydml
@@ -435,7 +435,7 @@ if (is_supported == 1):
save (beta_out, fileB, format=fmtB)
if (intercept_status == 1 | intercept_status == 2):
- intercept_value = castAsScalar (beta_out [num_features, 1])
+ intercept_value = scalar (beta_out [num_features, 1])
beta_noicept = beta_out [1 : (num_features - 1), 1]
else:
beta_noicept = beta_out [1 : num_features, 1]
@@ -443,9 +443,9 @@ if (is_supported == 1):
min_beta = min (beta_noicept)
max_beta = max (beta_noicept)
tmp_i_min_beta = rowIndexMin (transpose(beta_noicept))
- i_min_beta = castAsScalar (tmp_i_min_beta [1, 1])
+ i_min_beta = scalar (tmp_i_min_beta [1, 1])
tmp_i_max_beta = rowIndexMax (transpose(beta_noicept))
- i_max_beta = castAsScalar (tmp_i_max_beta [1, 1])
+ i_max_beta = scalar (tmp_i_max_beta [1, 1])
##### OVER-DISPERSION PART #####
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/id3/id3.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/id3/id3.dml b/src/test/scripts/applications/id3/id3.dml
index e033ca2..a127fc8 100644
--- a/src/test/scripts/applications/id3/id3.dml
+++ b/src/test/scripts/applications/id3/id3.dml
@@ -107,7 +107,7 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
num_samples = sum(X_subset)
print("num non zero labels: " + num_non_zero_labels)
- mpl = castAsScalar(most_popular_label)
+ mpl = as.scalar(most_popular_label)
print("most popular label: " + mpl)
print("num remaining attrs: " + num_remaining_attrs)
@@ -135,14 +135,14 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
sz = nrow(attributes)
gains = matrix(0, rows=sz, cols=1)
for(i in 1:nrow(attributes)){
- if(castAsScalar(attributes[i,1]) == 1){
+ if(as.scalar(attributes[i,1]) == 1){
attr_vals = X[,i]
attr_domain = aggregate(target=X_subset, groups=attr_vals, fn="sum")
hxt_vector = matrix(0, rows=nrow(attr_domain), cols=1)
for(j in 1:nrow(attr_domain), check=0){
- if(castAsScalar(attr_domain[j,1]) != 0){
+ if(as.scalar(attr_domain[j,1]) != 0){
val = j
Tj = X_subset * ppred(X[,i], val, "==")
@@ -168,8 +168,8 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
max_gain = 0
for(i4 in 1:nrow(gains)){
#print("best attr " + best_attr + " max gain " + max_gain)
- if(castAsScalar(attributes[i4,1]) == 1){
- g = castAsScalar(gains[i4,1])
+ if(as.scalar(attributes[i4,1]) == 1){
+ g = as.scalar(gains[i4,1])
if(best_attr == -1 | max_gain <= g){
max_gain = g
best_attr = i4
@@ -212,7 +212,7 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
start_pt = 1+(i1-1)*max_sz_subtree
tempNodeStore[,start_pt:(start_pt+nrow(nodesi)-1)] = t(nodesi)
numSubtreeNodes[i1,1] = nrow(nodesi)
- if(nrow(edgesi)!=1 | ncol(edgesi)!=1 | castAsScalar(edgesi[1,1])!=-1){
+ if(nrow(edgesi)!=1 | ncol(edgesi)!=1 | as.scalar(edgesi[1,1])!=-1){
tempEdgeStore[,start_pt:(start_pt+nrow(edgesi)-1)] = t(edgesi)
numSubtreeEdges[i1,1] = nrow(edgesi)
}else{
@@ -239,7 +239,7 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
edges = matrix(1, rows=sz, cols=3)
numEdges = 0
for(i6 in 1:nrow(attr_domain)){
- num_nodesi = castAsScalar(numSubtreeNodes[i6,1])
+ num_nodesi = as.scalar(numSubtreeNodes[i6,1])
if(num_nodesi > 0){
edges[numEdges+1,2] = i6
numEdges = numEdges + 1
@@ -248,13 +248,13 @@ id3_learn = function(Matrix[Double] X, Matrix[Double] y, Matrix[Double] X_subset
nonEmptyAttri = 0
for(i7 in 1:nrow(attr_domain)){
- numNodesInSubtree = castAsScalar(numSubtreeNodes[i7,1])
+ numNodesInSubtree = as.scalar(numSubtreeNodes[i7,1])
if(numNodesInSubtree > 0){
start_pt1 = 1 + (i7-1)*max_sz_subtree
nodes[numNodes+1:numNodes+numNodesInSubtree,] = t(tempNodeStore[,start_pt1:(start_pt1+numNodesInSubtree-1)])
- numEdgesInSubtree = castAsScalar(numSubtreeEdges[i7,1])
+ numEdgesInSubtree = as.scalar(numSubtreeEdges[i7,1])
if(numEdgesInSubtree!=0){
edgesi1 = t(tempEdgeStore[,start_pt1:(start_pt1+numEdgesInSubtree-1)])
@@ -298,9 +298,9 @@ y = y + labelCorrection + 0
nodes[,2] = nodes[,2] - labelCorrection * ppred(nodes[,1], -1, "==")
for(i3 in 1:nrow(edges)){
#parfor(i3 in 1:nrow(edges)){
- e_parent = castAsScalar(edges[i3,1])
- parent_feature = castAsScalar(nodes[e_parent,1])
- correction = castAsScalar(featureCorrections[1,parent_feature])
+ e_parent = as.scalar(edges[i3,1])
+ parent_feature = as.scalar(nodes[e_parent,1])
+ correction = as.scalar(featureCorrections[1,parent_feature])
edges[i3,2] = edges[i3,2] - correction
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/id3/id3.pydml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/id3/id3.pydml b/src/test/scripts/applications/id3/id3.pydml
index a59e901..17a10e4 100644
--- a/src/test/scripts/applications/id3/id3.pydml
+++ b/src/test/scripts/applications/id3/id3.pydml
@@ -107,7 +107,7 @@ def id3_learn(X:matrix[float], y:matrix[float], X_subset:matrix[float], attribut
num_samples = sum(X_subset)
print("num non zero labels: " + num_non_zero_labels)
- mpl = castAsScalar(most_popular_label)
+ mpl = scalar(most_popular_label)
print("most popular label: " + mpl)
print("num remaining attrs: " + num_remaining_attrs)
@@ -135,14 +135,14 @@ def id3_learn(X:matrix[float], y:matrix[float], X_subset:matrix[float], attribut
sz = nrow(attributes)
gains = full(0, rows=sz, cols=1)
for(i in 1:nrow(attributes)):
- if(castAsScalar(attributes[i,1]) == 1):
+ if(scalar(attributes[i,1]) == 1):
attr_vals = X[,i]
attr_domain = aggregate(target=X_subset, groups=attr_vals, fn="sum")
hxt_vector = full(0, rows=nrow(attr_domain), cols=1)
for(j in 1:nrow(attr_domain), check=0):
- if(castAsScalar(attr_domain[j,1]) != 0):
+ if(scalar(attr_domain[j,1]) != 0):
val = j
Tj = X_subset * ppred(X[,i], val, "==")
@@ -164,8 +164,8 @@ def id3_learn(X:matrix[float], y:matrix[float], X_subset:matrix[float], attribut
max_gain = 0
for(i4 in 1:nrow(gains)):
#print("best attr " + best_attr + " max gain " + max_gain)
- if(castAsScalar(attributes[i4,1]) == 1):
- g = castAsScalar(gains[i4,1])
+ if(scalar(attributes[i4,1]) == 1):
+ g = scalar(gains[i4,1])
if(best_attr == -1 | max_gain <= g):
max_gain = g
best_attr = i4
@@ -203,7 +203,7 @@ def id3_learn(X:matrix[float], y:matrix[float], X_subset:matrix[float], attribut
start_pt = 1+(i1-1)*max_sz_subtree
tempNodeStore[,start_pt:(start_pt+nrow(nodesi)-1)] = t(nodesi)
numSubtreeNodes[i1,1] = nrow(nodesi)
- if(nrow(edgesi)!=1 | ncol(edgesi)!=1 | castAsScalar(edgesi[1,1])!=-1):
+ if(nrow(edgesi)!=1 | ncol(edgesi)!=1 | scalar(edgesi[1,1])!=-1):
tempEdgeStore[,start_pt:(start_pt+nrow(edgesi)-1)] = t(edgesi)
numSubtreeEdges[i1,1] = nrow(edgesi)
else:
@@ -227,20 +227,20 @@ def id3_learn(X:matrix[float], y:matrix[float], X_subset:matrix[float], attribut
edges = full(1, rows=sz, cols=3)
numEdges = 0
for(i6 in 1:nrow(attr_domain)):
- num_nodesi = castAsScalar(numSubtreeNodes[i6,1])
+ num_nodesi = scalar(numSubtreeNodes[i6,1])
if(num_nodesi > 0):
edges[numEdges+1,2] = i6
numEdges = numEdges + 1
nonEmptyAttri = 0
for(i7 in 1:nrow(attr_domain)):
- numNodesInSubtree = castAsScalar(numSubtreeNodes[i7,1])
+ numNodesInSubtree = scalar(numSubtreeNodes[i7,1])
if(numNodesInSubtree > 0):
start_pt1 = 1 + (i7-1)*max_sz_subtree
nodes[numNodes+1:numNodes+numNodesInSubtree,] = transpose(tempNodeStore[,start_pt1:(start_pt1+numNodesInSubtree-1)])
- numEdgesInSubtree = castAsScalar(numSubtreeEdges[i7,1])
+ numEdgesInSubtree = scalar(numSubtreeEdges[i7,1])
if(numEdgesInSubtree!=0):
edgesi1 = transpose(tempEdgeStore[,start_pt1:(start_pt1+numEdgesInSubtree-1)])
@@ -279,9 +279,9 @@ y = y + labelCorrection + 0
nodes[,2] = nodes[,2] - labelCorrection * ppred(nodes[,1], -1, "==")
for(i3 in 1:nrow(edges)):
#parfor(i3 in 1:nrow(edges)):
- e_parent = castAsScalar(edges[i3,1])
- parent_feature = castAsScalar(nodes[e_parent,1])
- correction = castAsScalar(featureCorrections[1,parent_feature])
+ e_parent = scalar(edges[i3,1])
+ parent_feature = scalar(nodes[e_parent,1])
+ correction = scalar(featureCorrections[1,parent_feature])
edges[i3,2] = edges[i3,2] - correction
save(nodes, $3, format="text")
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/imputeGaussMCMC.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/imputeGaussMCMC.dml b/src/test/scripts/applications/impute/imputeGaussMCMC.dml
index 21ecaee..6575d3b 100644
--- a/src/test/scripts/applications/impute/imputeGaussMCMC.dml
+++ b/src/test/scripts/applications/impute/imputeGaussMCMC.dml
@@ -179,7 +179,7 @@ while (is_opt_converged == 0)
}
q = t(gradient_change_p);
- alpha = norm_r2 / castAsScalar (t(p) %*% q);
+ alpha = norm_r2 / as.scalar (t(p) %*% q);
shift_vector_change = alpha * p;
shift_vector = shift_vector + shift_vector_change;
old_norm_r2 = norm_r2;
@@ -286,7 +286,7 @@ while (is_enough_gradient_descent == 0)
q_frees = t(gradientInFrees_eps_p - gradientInFrees) / cg_eps;
q_params = t(gradientInParams_eps_p - gradientInParams) / cg_eps;
- alpha = norm_r2 / castAsScalar (t(p_frees) %*% q_frees + t(p_params) %*% q_params);
+ alpha = norm_r2 / as.scalar (t(p_frees) %*% q_frees + t(p_params) %*% q_params);
shift_frees = shift_frees + alpha * p_frees;
shift_params = shift_params + alpha * p_params;
@@ -475,8 +475,8 @@ left_swap = round (0.5 + dim_sample * rnd);
rnd = Rand (rows = num_swaps, cols = 1, min = 0.0, max = 1.0);
right_swap = round (0.5 + dim_sample * rnd);
for (swap_i in 1:num_swaps) {
- l = castAsScalar (left_swap [swap_i, 1]);
- r = castAsScalar (right_swap [swap_i, 1]);
+ l = as.scalar (left_swap [swap_i, 1]);
+ r = as.scalar (right_swap [swap_i, 1]);
if (l != r) {
tmp_row = SampleOrder [l, ];
SampleOrder [l, ] = SampleOrder [r, ];
@@ -560,7 +560,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
# Create a normally distributed random sample
- dim_half_sample = castAsScalar (round (dim_sample / 2 + 0.1 + zero));
+ dim_half_sample = as.scalar (round (dim_sample / 2 + 0.1 + zero));
rnd1 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd2 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd_normal_1 = sqrt (- 2.0 * log (rnd1)) * sin (2 * pi * rnd2);
@@ -583,7 +583,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
{
# Generate the sample unit-vector and updaters
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars_updater = SampleOrder [1 : num_frees, idx];
regresValues_updater = RegresValueMap %*% CReps %*% freeVars_updater;
bilinear_updater_vector = regresValues_updater * regresParams;
@@ -607,18 +607,18 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
mean_shift = - coeff_b / (2.0 * coeff_a);
sigma_shift = 1.0 / sqrt (2.0 * coeff_a);
- shift = mean_shift + sigma_shift * castAsScalar (rnd_normal [idx, 1]);
+ shift = mean_shift + sigma_shift * as.scalar (rnd_normal [idx, 1]);
# BEGIN DEBUG INSERT
# mmm = 1;
-# if (castAsScalar (isVar [1, idx]) > 0.5 & # IT IS A FREE VARIABLE, NOT A PARAMETER
-# castAsScalar (freeVars_updater [mmm, 1]) > 0) # IT IS mmm-TH FREE VARIABLE
+# if (as.scalar (isVar [1, idx]) > 0.5 & # IT IS A FREE VARIABLE, NOT A PARAMETER
+# as.scalar (freeVars_updater [mmm, 1]) > 0) # IT IS mmm-TH FREE VARIABLE
# {
# # print ("freeVars[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a + ", coeff_b = " + coeff_b);
# print ("freeVars[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", mean_shift = " + mean_shift + ", sigma_shift = " + sigma_shift + ", shift = " + shift);
# }
-# if (castAsScalar (isVar [1, idx]) <= 0.5 & # IT IS A PARAMETER, NOT A FREE VARIABLE
-# castAsScalar (params_updater [mmm, 1]) > 0) # IT IS mmm-TH PARAMETER
+# if (as.scalar (isVar [1, idx]) <= 0.5 & # IT IS A PARAMETER, NOT A FREE VARIABLE
+# as.scalar (params_updater [mmm, 1]) > 0) # IT IS mmm-TH PARAMETER
# {
# # print (" params[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a + ", coeff_b = " + coeff_b);
# print (" params[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", mean_shift = " + mean_shift + ", sigma_shift = " + sigma_shift + ", shift = " + shift);
@@ -628,7 +628,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
# Perform the updates
bilinear_form = bilinear_form + shift * bilinear_updater;
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars = freeVars + shift * freeVars_updater;
regresValues = regresValues + shift * regresValues_updater;
} else {
@@ -653,21 +653,21 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
sum_of_observed_losses = sum_of_observed_losses + bilinear_form_value;
}
-# v1 =castAsScalar(round(10000*reports[1 + (num_terms - 1) * num_attrs, 1])/10000);
-# v2 =castAsScalar(round(10000*reports[2 + (num_terms - 1) * num_attrs, 1])/10000);
-# v3 =castAsScalar(round(10000*reports[3 + (num_terms - 1) * num_attrs, 1])/10000);
-# v4 =castAsScalar(round(10000*reports[4 + (num_terms - 1) * num_attrs, 1])/10000);
-# w1 =castAsScalar(round(10000*reports_matrix[ 1,num_terms])/10000);
-# w2 =castAsScalar(round(10000*reports_matrix[ 2,num_terms])/10000);
-# w3 =castAsScalar(round(10000*reports_matrix[ 3,num_terms])/10000);
-# w4 =castAsScalar(round(10000*reports_matrix[ 4,num_terms])/10000);
-
-# v5 =castAsScalar(round(reports_matrix[ 5,num_terms]));
-# v8 =castAsScalar(round(reports_matrix[ 8,num_terms]));
-# v9 =castAsScalar(round(reports_matrix[ 9,num_terms]));
-# v10=castAsScalar(round(reports_matrix[10,num_terms]));
-# v16=castAsScalar(round(reports_matrix[16,num_terms]));
-# v19=castAsScalar(round(reports_matrix[19,num_terms]));
+# v1 =as.scalar(round(10000*reports[1 + (num_terms - 1) * num_attrs, 1])/10000);
+# v2 =as.scalar(round(10000*reports[2 + (num_terms - 1) * num_attrs, 1])/10000);
+# v3 =as.scalar(round(10000*reports[3 + (num_terms - 1) * num_attrs, 1])/10000);
+# v4 =as.scalar(round(10000*reports[4 + (num_terms - 1) * num_attrs, 1])/10000);
+# w1 =as.scalar(round(10000*reports_matrix[ 1,num_terms])/10000);
+# w2 =as.scalar(round(10000*reports_matrix[ 2,num_terms])/10000);
+# w3 =as.scalar(round(10000*reports_matrix[ 3,num_terms])/10000);
+# w4 =as.scalar(round(10000*reports_matrix[ 4,num_terms])/10000);
+
+# v5 =as.scalar(round(reports_matrix[ 5,num_terms]));
+# v8 =as.scalar(round(reports_matrix[ 8,num_terms]));
+# v9 =as.scalar(round(reports_matrix[ 9,num_terms]));
+# v10=as.scalar(round(reports_matrix[10,num_terms]));
+# v16=as.scalar(round(reports_matrix[16,num_terms]));
+# v19=as.scalar(round(reports_matrix[19,num_terms]));
#print (" Sample = 1:" + v1 + ", 2:" + v2 + ", 3:" + v3 + ", 4:" + v4);
## + ", 5:" + v5 + ", 8:" + v8 + ", 9:" + v9 + ", 10:" + v10 + ", 16:" + v16 + ", 19:" + v19);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/imputeGaussMCMC.nogradient.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/imputeGaussMCMC.nogradient.dml b/src/test/scripts/applications/impute/imputeGaussMCMC.nogradient.dml
index 00210c5..897fc21 100644
--- a/src/test/scripts/applications/impute/imputeGaussMCMC.nogradient.dml
+++ b/src/test/scripts/applications/impute/imputeGaussMCMC.nogradient.dml
@@ -181,7 +181,7 @@ while (is_opt_converged == 0)
q [i, 1] = (quadratic_plus_both - quadratic_plus_1 - quadratic_plus_p + bilinear_form_value) + q [i, 1];
}
- alpha = norm_r2 / castAsScalar (t(p) %*% q);
+ alpha = norm_r2 / as.scalar (t(p) %*% q);
shift_vector = shift_vector + alpha * p;
old_norm_r2 = norm_r2;
residual = residual + alpha * q;
@@ -238,8 +238,8 @@ left_swap = round (0.5 + dim_sample * rnd);
rnd = Rand (rows = num_swaps, cols = 1, min = 0.0, max = 1.0);
right_swap = round (0.5 + dim_sample * rnd);
for (swap_i in 1:num_swaps) {
- l = castAsScalar (left_swap [swap_i, 1]);
- r = castAsScalar (right_swap [swap_i, 1]);
+ l = as.scalar (left_swap [swap_i, 1]);
+ r = as.scalar (right_swap [swap_i, 1]);
if (l != r) {
tmp_row = SampleOrder [l, ];
SampleOrder [l, ] = SampleOrder [r, ];
@@ -324,7 +324,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
# Create a normally distributed random sample
- dim_half_sample = castAsScalar (round (dim_sample / 2 + 0.1 + zero));
+ dim_half_sample = as.scalar (round (dim_sample / 2 + 0.1 + zero));
rnd1 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd2 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd_normal_1 = sqrt (- 2.0 * log (rnd1)) * sin (2 * pi * rnd2);
@@ -347,7 +347,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
{
# Generate the sample unit-vector and updaters
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars_updater = SampleOrder [1 : num_frees, idx];
regresValues_updater = RegresValueMap %*% CReps %*% freeVars_updater;
bilinear_updater_vector = regresValues_updater * regresParams;
@@ -372,18 +372,18 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
mean_shift = - coeff_b / (2.0 * coeff_a);
sigma_shift = 1.0 / sqrt (2.0 * coeff_a);
- shift = mean_shift + sigma_shift * castAsScalar (rnd_normal [idx, 1]);
+ shift = mean_shift + sigma_shift * as.scalar (rnd_normal [idx, 1]);
# BEGIN DEBUG INSERT
# mmm = 1;
-# if (castAsScalar (isVar [1, idx]) > 0.5 & # IT IS A FREE VARIABLE, NOT A PARAMETER
-# castAsScalar (freeVars_updater [mmm, 1]) > 0) # IT IS mmm-TH FREE VARIABLE
+# if (as.scalar (isVar [1, idx]) > 0.5 & # IT IS A FREE VARIABLE, NOT A PARAMETER
+# as.scalar (freeVars_updater [mmm, 1]) > 0) # IT IS mmm-TH FREE VARIABLE
# {
# # print ("freeVars[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a + ", coeff_b = " + coeff_b);
# print ("freeVars[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", mean_shift = " + mean_shift + ", sigma_shift = " + sigma_shift + ", shift = " + shift);
# }
-# if (castAsScalar (isVar [1, idx]) <= 0.5 & # IT IS A PARAMETER, NOT A FREE VARIABLE
-# castAsScalar (params_updater [mmm, 1]) > 0) # IT IS mmm-TH PARAMETER
+# if (as.scalar (isVar [1, idx]) <= 0.5 & # IT IS A PARAMETER, NOT A FREE VARIABLE
+# as.scalar (params_updater [mmm, 1]) > 0) # IT IS mmm-TH PARAMETER
# {
# # print (" params[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a + ", coeff_b = " + coeff_b);
# print (" params[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", mean_shift = " + mean_shift + ", sigma_shift = " + sigma_shift + ", shift = " + shift);
@@ -393,7 +393,7 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
# Perform the updates
bilinear_form = bilinear_form + shift * bilinear_updater;
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars = freeVars + shift * freeVars_updater;
regresValues = regresValues + shift * regresValues_updater;
} else {
@@ -419,21 +419,21 @@ while ((iter < max_num_iter) & (num_of_observed_reports < max_num_observed_itera
sum_of_observed_losses = sum_of_observed_losses + bilinear_form_value;
}
-# v1 =castAsScalar(round(10000*reports[1 + (num_terms - 1) * num_attrs, 1])/10000);
-# v2 =castAsScalar(round(10000*reports[2 + (num_terms - 1) * num_attrs, 1])/10000);
-# v3 =castAsScalar(round(10000*reports[3 + (num_terms - 1) * num_attrs, 1])/10000);
-# v4 =castAsScalar(round(10000*reports[4 + (num_terms - 1) * num_attrs, 1])/10000);
-# w1 =castAsScalar(round(10000*reports_matrix[ 1,num_terms])/10000);
-# w2 =castAsScalar(round(10000*reports_matrix[ 2,num_terms])/10000);
-# w3 =castAsScalar(round(10000*reports_matrix[ 3,num_terms])/10000);
-# w4 =castAsScalar(round(10000*reports_matrix[ 4,num_terms])/10000);
-
-# v5 =castAsScalar(round(reports_matrix[ 5,num_terms]));
-# v8 =castAsScalar(round(reports_matrix[ 8,num_terms]));
-# v9 =castAsScalar(round(reports_matrix[ 9,num_terms]));
-# v10=castAsScalar(round(reports_matrix[10,num_terms]));
-# v16=castAsScalar(round(reports_matrix[16,num_terms]));
-# v19=castAsScalar(round(reports_matrix[19,num_terms]));
+# v1 =as.scalar(round(10000*reports[1 + (num_terms - 1) * num_attrs, 1])/10000);
+# v2 =as.scalar(round(10000*reports[2 + (num_terms - 1) * num_attrs, 1])/10000);
+# v3 =as.scalar(round(10000*reports[3 + (num_terms - 1) * num_attrs, 1])/10000);
+# v4 =as.scalar(round(10000*reports[4 + (num_terms - 1) * num_attrs, 1])/10000);
+# w1 =as.scalar(round(10000*reports_matrix[ 1,num_terms])/10000);
+# w2 =as.scalar(round(10000*reports_matrix[ 2,num_terms])/10000);
+# w3 =as.scalar(round(10000*reports_matrix[ 3,num_terms])/10000);
+# w4 =as.scalar(round(10000*reports_matrix[ 4,num_terms])/10000);
+
+# v5 =as.scalar(round(reports_matrix[ 5,num_terms]));
+# v8 =as.scalar(round(reports_matrix[ 8,num_terms]));
+# v9 =as.scalar(round(reports_matrix[ 9,num_terms]));
+# v10=as.scalar(round(reports_matrix[10,num_terms]));
+# v16=as.scalar(round(reports_matrix[16,num_terms]));
+# v19=as.scalar(round(reports_matrix[19,num_terms]));
#print (" Sample = 1:" + v1 + ", 2:" + v2 + ", 3:" + v3 + ", 4:" + v4);
## + ", 5:" + v5 + ", 8:" + v8 + ", 9:" + v9 + ", 10:" + v10 + ", 16:" + v16 + ", 19:" + v19);
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/old/imputeGaussMCMC.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/old/imputeGaussMCMC.dml b/src/test/scripts/applications/impute/old/imputeGaussMCMC.dml
index 77bd21c..7f9a875 100644
--- a/src/test/scripts/applications/impute/old/imputeGaussMCMC.dml
+++ b/src/test/scripts/applications/impute/old/imputeGaussMCMC.dml
@@ -173,7 +173,7 @@ while (is_opt_converged == 0)
q [i, 1] = (quadratic_plus_both - quadratic_plus_1 - quadratic_plus_p + bilinear_form_value) + q [i, 1];
}
- alpha = norm_r2 / castAsScalar (t(p) %*% q);
+ alpha = norm_r2 / as.scalar (t(p) %*% q);
shift_vector = shift_vector + alpha * p;
old_norm_r2 = norm_r2;
residual = residual + alpha * q;
@@ -228,8 +228,8 @@ left_swap = round (0.5 + dim_sample * rnd);
rnd = Rand (rows = num_swaps, cols = 1, min = 0.0, max = 1.0);
right_swap = round (0.5 + dim_sample * rnd);
for (swap_i in 1:num_swaps) {
- l = castAsScalar (left_swap [swap_i, 1]);
- r = castAsScalar (right_swap [swap_i, 1]);
+ l = as.scalar (left_swap [swap_i, 1]);
+ r = as.scalar (right_swap [swap_i, 1]);
if (l != r) {
tmp_row = SampleOrder [l, ];
SampleOrder [l, ] = SampleOrder [r, ];
@@ -265,7 +265,7 @@ for (iter in 1:num_iter)
# Create a normally distributed random sample
- dim_half_sample = castAsScalar (round (dim_sample / 2 + 0.1 + zero));
+ dim_half_sample = as.scalar (round (dim_sample / 2 + 0.1 + zero));
rnd1 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd2 = Rand (rows = dim_half_sample, cols = 1, min = 0.0, max = 1.0);
rnd_normal_1 = sqrt (- 2.0 * log (rnd1)) * sin (2 * pi * rnd2);
@@ -288,7 +288,7 @@ for (iter in 1:num_iter)
{
# Generate the sample unit-vector and updaters
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars_updater = SampleOrder [1 : num_frees, idx];
regresValues_updater = RegresValueMap %*% CReps %*% freeVars_updater;
bilinear_updater_vector = regresValues_updater * regresParams;
@@ -309,11 +309,11 @@ for (iter in 1:num_iter)
# BEGIN DEBUG INSERT
# mmm = 1;
-# if (castAsScalar (isVar [1, idx]) > 0.5) {
-# for (iii in 2:num_frees) {if (castAsScalar (freeVars_updater [iii, 1] - freeVars_updater [mmm, 1]) > 0) {mmm = iii;}}
+# if (as.scalar (isVar [1, idx]) > 0.5) {
+# for (iii in 2:num_frees) {if (as.scalar (freeVars_updater [iii, 1] - freeVars_updater [mmm, 1]) > 0) {mmm = iii;}}
# print ("freeVars[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a);
# } else {
-# for (iii in 2:num_params) {if (castAsScalar (params_updater [iii, 1] - params_updater [mmm, 1]) > 0) {mmm = iii;}}
+# for (iii in 2:num_params) {if (as.scalar (params_updater [iii, 1] - params_updater [mmm, 1]) > 0) {mmm = iii;}}
# print (" params[" + mmm + "]: q_minus_1 = " + q_minus_1 + ", q_plus_1 = " + q_plus_1 + ", coeff_a = " + coeff_a);
# }
# END DEBUG INSERT
@@ -323,12 +323,12 @@ for (iter in 1:num_iter)
mean_shift = - coeff_b / (2.0 * coeff_a);
sigma_shift = 1.0 / sqrt (2.0 * coeff_a);
- shift = mean_shift + sigma_shift * castAsScalar (rnd_normal [idx, 1]);
+ shift = mean_shift + sigma_shift * as.scalar (rnd_normal [idx, 1]);
# Perform the updates
bilinear_form = bilinear_form + shift * bilinear_updater;
- if (castAsScalar (isVar [1, idx]) > 0.5) {
+ if (as.scalar (isVar [1, idx]) > 0.5) {
freeVars = freeVars + shift * freeVars_updater;
regresValues = regresValues + shift * regresValues_updater;
} else {
@@ -353,16 +353,16 @@ for (iter in 1:num_iter)
}
-v1 =castAsScalar(round(reports_matrix[ 1,num_terms]));
-v2 =castAsScalar(round(reports_matrix[ 2,num_terms]));
-v3 =castAsScalar(round(reports_matrix[ 3,num_terms]));
-v4 =castAsScalar(round(reports_matrix[ 4,num_terms]));
-v5 =castAsScalar(round(reports_matrix[ 5,num_terms]));
-v8 =castAsScalar(round(reports_matrix[ 8,num_terms]));
-v9 =castAsScalar(round(reports_matrix[ 9,num_terms]));
-v10=castAsScalar(round(reports_matrix[10,num_terms]));
-v16=castAsScalar(round(reports_matrix[16,num_terms]));
-v19=castAsScalar(round(reports_matrix[19,num_terms]));
+v1 =as.scalar(round(reports_matrix[ 1,num_terms]));
+v2 =as.scalar(round(reports_matrix[ 2,num_terms]));
+v3 =as.scalar(round(reports_matrix[ 3,num_terms]));
+v4 =as.scalar(round(reports_matrix[ 4,num_terms]));
+v5 =as.scalar(round(reports_matrix[ 5,num_terms]));
+v8 =as.scalar(round(reports_matrix[ 8,num_terms]));
+v9 =as.scalar(round(reports_matrix[ 9,num_terms]));
+v10=as.scalar(round(reports_matrix[10,num_terms]));
+v16=as.scalar(round(reports_matrix[16,num_terms]));
+v19=as.scalar(round(reports_matrix[19,num_terms]));
print (
" Sample = 1:" + v1 + ", 2:" + v2 + ", 3:" + v3 + ", 4:" + v4 + ", 5:" + v5 +
", 8:" + v8 + ", 9:" + v9 + ", 10:" + v10 + ", 16:" + v16 + ", 19:" + v19);
@@ -402,7 +402,7 @@ matricize = function (Matrix[double] v, int n_rows) return (Matrix[double] M)
{
zero = matrix (0.0, rows = 1, cols = 1);
n = nrow (v);
- n_cols = castAsScalar (round (zero + (n / n_rows)));
+ n_cols = as.scalar (round (zero + (n / n_rows)));
if (n_cols * n_rows < n) {
n_cols = n_cols + 1;
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/tmp.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/tmp.dml b/src/test/scripts/applications/impute/tmp.dml
index 9e0417e..c852cce 100644
--- a/src/test/scripts/applications/impute/tmp.dml
+++ b/src/test/scripts/applications/impute/tmp.dml
@@ -26,7 +26,7 @@ blahblah = 0.0 / 0.0; # -0.00099999999;
print (blahblah);
x = matrix (0.0, rows = 55, cols = 1);
x [55, 1] = blahblah;
-print (castAsScalar (x [55, 1]));
+print (as.scalar (x [55, 1]));
for (i in 1:9) {
x [i, 1] = -0.001 * i;
}
@@ -36,9 +36,9 @@ for (i in 1:5) {
y = atan_temporary (x);
z = tan (y);
for (i in 1:nrow(x)) {
- [x_m, x_e] = round_to_print (castAsScalar (x[i,1]));
- [a_m, a_e] = round_to_print (castAsScalar (y[i,1]));
- [t_m, t_e] = round_to_print (castAsScalar (z[i,1]));
+ [x_m, x_e] = round_to_print (as.scalar (x[i,1]));
+ [a_m, a_e] = round_to_print (as.scalar (y[i,1]));
+ [t_m, t_e] = round_to_print (as.scalar (z[i,1]));
print ("x = " + x_m + "E" + x_e + "; atan(x) = " + a_m + "E" + a_e + "; tan(atan(x)) = " + t_m + "E" + t_e);
}
http://git-wip-us.apache.org/repos/asf/incubator-systemml/blob/2da81457/src/test/scripts/applications/impute/wfundInputGenerator1.dml
----------------------------------------------------------------------
diff --git a/src/test/scripts/applications/impute/wfundInputGenerator1.dml b/src/test/scripts/applications/impute/wfundInputGenerator1.dml
index 8457fbd..5f7824b 100644
--- a/src/test/scripts/applications/impute/wfundInputGenerator1.dml
+++ b/src/test/scripts/applications/impute/wfundInputGenerator1.dml
@@ -380,7 +380,7 @@ if (is_GROUP_4_ENABLED == 1) {
for (t in (num_known_terms + num_predicted_terms + 1) : num_terms)
{
for (i in 1 : num_attrs) {
- if (castAsScalar (disabled_known_values [i, t - (num_known_terms + num_predicted_terms)]) == 0.0)
+ if (as.scalar (disabled_known_values [i, t - (num_known_terms + num_predicted_terms)]) == 0.0)
{
reg_index = ((t-1) * num_attrs - 1 + i) * num_factors;
RegresCoeffDefault [reg_index + 1, 1] = 1.0 + zero; # Default coefficient = 1.0