You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@mahout.apache.org by gs...@apache.org on 2009/12/18 00:22:41 UTC
svn commit: r891983 [27/47] - in /lucene/mahout/trunk: ./ core/
core/src/main/java/org/apache/mahout/cf/taste/hadoop/item/
core/src/main/java/org/apache/mahout/clustering/
core/src/main/java/org/apache/mahout/clustering/canopy/
core/src/main/java/org/a...
Added: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractIntObjectMap.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractIntObjectMap.java?rev=891983&view=auto
==============================================================================
--- lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractIntObjectMap.java (added)
+++ lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractIntObjectMap.java Thu Dec 17 23:22:16 2009
@@ -0,0 +1,442 @@
+/*
+Copyright � 1999 CERN - European Organization for Nuclear Research.
+Permission to use, copy, modify, distribute and sell this software and its documentation for any purpose
+is hereby granted without fee, provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear in supporting documentation.
+CERN makes no representations about the suitability of this software for any purpose.
+It is provided "as is" without expressed or implied warranty.
+*/
+package org.apache.mahout.math.map;
+
+import org.apache.mahout.math.GenericSorting;
+import org.apache.mahout.math.Swapper;
+import org.apache.mahout.math.function.IntComparator;
+import org.apache.mahout.math.function.IntObjectProcedure;
+import org.apache.mahout.math.function.IntProcedure;
+import org.apache.mahout.math.list.IntArrayList;
+import org.apache.mahout.math.list.ObjectArrayList;
+
+/** @deprecated until unit tests are in place. Until this time, this class/interface is unsupported. */
+@Deprecated
+public abstract class AbstractIntObjectMap extends AbstractMap {
+ //public static int hashCollisions = 0; // for debug only
+
+ /** Makes this class non instantiable, but still let's others inherit from it. */
+ protected AbstractIntObjectMap() {
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified key.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified key.
+ */
+ public boolean containsKey(final int key) {
+ return !forEachKey(
+ new IntProcedure() {
+ @Override
+ public boolean apply(int iterKey) {
+ return (key != iterKey);
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified value. Tests for identity.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified value.
+ */
+ public boolean containsValue(final Object value) {
+ return !forEachPair(
+ new IntObjectProcedure() {
+ @Override
+ public boolean apply(int iterKey, Object iterValue) {
+ return (value != iterValue);
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns a deep copy of the receiver; uses <code>clone()</code> and casts the result.
+ *
+ * @return a deep copy of the receiver.
+ */
+ public AbstractIntObjectMap copy() {
+ return (AbstractIntObjectMap) clone();
+ }
+
+ /**
+ * Compares the specified object with this map for equality. Returns <tt>true</tt> if the given object is also a map
+ * and the two maps represent the same mappings. More formally, two maps <tt>m1</tt> and <tt>m2</tt> represent the
+ * same mappings iff
+ * <pre>
+ * m1.forEachPair(
+ * new IntObjectProcedure() {
+ * public boolean apply(int key, Object value) {
+ * return m2.containsKey(key) && m2.get(key) == value;
+ * }
+ * }
+ * )
+ * &&
+ * m2.forEachPair(
+ * new IntObjectProcedure() {
+ * public boolean apply(int key, Object value) {
+ * return m1.containsKey(key) && m1.get(key) == value;
+ * }
+ * }
+ * );
+ * </pre>
+ *
+ * This implementation first checks if the specified object is this map; if so it returns <tt>true</tt>. Then, it
+ * checks if the specified object is a map whose size is identical to the size of this set; if not, it it returns
+ * <tt>false</tt>. If so, it applies the iteration as described above.
+ *
+ * @param obj object to be compared for equality with this map.
+ * @return <tt>true</tt> if the specified object is equal to this map.
+ */
+ public boolean equals(Object obj) {
+ if (obj == this) {
+ return true;
+ }
+
+ if (!(obj instanceof AbstractIntObjectMap)) {
+ return false;
+ }
+ final AbstractIntObjectMap other = (AbstractIntObjectMap) obj;
+ if (other.size() != size()) {
+ return false;
+ }
+
+ return
+ forEachPair(
+ new IntObjectProcedure() {
+ @Override
+ public boolean apply(int key, Object value) {
+ return other.containsKey(key) && other.get(key) == value;
+ }
+ }
+ )
+ &&
+ other.forEachPair(
+ new IntObjectProcedure() {
+ @Override
+ public boolean apply(int key, Object value) {
+ return containsKey(key) && get(key) == value;
+ }
+ }
+ );
+ }
+
+ /**
+ * Applies a procedure to each key of the receiver, if any. Note: Iterates over the keys in no particular order.
+ * Subclasses can define a particular order, for example, "sorted by key". All methods which <i>can</i> be expressed
+ * in terms of this method (most methods can) <i>must guarantee</i> to use the <i>same</i> order defined by this
+ * method, even if it is no particular order. This is necessary so that, for example, methods <tt>keys</tt> and
+ * <tt>values</tt> will yield association pairs, not two uncorrelated lists.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ public abstract boolean forEachKey(IntProcedure procedure);
+
+ /**
+ * Applies a procedure to each (key,value) pair of the receiver, if any. Iteration order is guaranteed to be
+ * <i>identical</i> to the order used by method {@link #forEachKey(IntProcedure)}.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ public boolean forEachPair(final IntObjectProcedure procedure) {
+ return forEachKey(
+ new IntProcedure() {
+ @Override
+ public boolean apply(int key) {
+ return procedure.apply(key, get(key));
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns the value associated with the specified key. It is often a good idea to first check with {@link
+ * #containsKey(int)} whether the given key has a value associated or not, i.e. whether there exists an association
+ * for the given key or not.
+ *
+ * @param key the key to be searched for.
+ * @return the value associated with the specified key; <tt>null</tt> if no such key is present.
+ */
+ public abstract Object get(int key);
+
+ /**
+ * Returns the first key the given value is associated with. It is often a good idea to first check with {@link
+ * #containsValue(Object)} whether there exists an association from a key to this value. Search order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(IntProcedure)}.
+ *
+ * @param value the value to search for.
+ * @return the first key for which holds <tt>get(key) == value</tt>; returns <tt>Integer.MIN_VALUE</tt> if no such key
+ * exists.
+ */
+ public int keyOf(final Object value) {
+ final int[] foundKey = new int[1];
+ boolean notFound = forEachPair(
+ new IntObjectProcedure() {
+ @Override
+ public boolean apply(int iterKey, Object iterValue) {
+ boolean found = value == iterValue;
+ if (found) {
+ foundKey[0] = iterKey;
+ }
+ return !found;
+ }
+ }
+ );
+ if (notFound) {
+ return Integer.MIN_VALUE;
+ }
+ return foundKey[0];
+ }
+
+ /**
+ * Returns a list filled with all keys contained in the receiver. The returned list has a size that equals
+ * <tt>this.size()</tt>. Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(IntProcedure)}. <p> This method can be used to iterate over the keys of the receiver.
+ *
+ * @return the keys.
+ */
+ public IntArrayList keys() {
+ IntArrayList list = new IntArrayList(size());
+ keys(list);
+ return list;
+ }
+
+ /**
+ * Fills all keys contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(IntProcedure)}. <p> This method can be used to
+ * iterate over the keys of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ public void keys(final IntArrayList list) {
+ list.clear();
+ forEachKey(
+ new IntProcedure() {
+ @Override
+ public boolean apply(int key) {
+ list.add(key);
+ return true;
+ }
+ }
+ );
+ }
+
+ /**
+ * Fills all keys <i>sorted ascending by their associated value</i> into the specified list. Fills into the list,
+ * starting at index 0. After this call returns the specified list has a new size that equals <tt>this.size()</tt>.
+ * Primary sort criterium is "value", secondary sort criterium is "key". This means that if any two values are equal,
+ * the smaller key comes first. <p> <b>Example:</b> <br> <tt>keys = (8,7,6), values = (1,2,2) --> keyList =
+ * (8,6,7)</tt>
+ *
+ * @param keyList the list to be filled, can have any size.
+ */
+ public void keysSortedByValue(IntArrayList keyList) {
+ pairsSortedByValue(keyList, new ObjectArrayList(size()));
+ }
+
+ /**
+ * Fills all pairs satisfying a given condition into the specified lists. Fills into the lists, starting at index 0.
+ * After this call returns the specified lists both have a new size, the number of pairs satisfying the condition.
+ * Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link #forEachKey(IntProcedure)}.
+ * <p> <b>Example:</b> <br>
+ * <pre>
+ * IntObjectProcedure condition = new IntObjectProcedure() { // match even keys only
+ * public boolean apply(int key, Object value) { return key%2==0; }
+ * }
+ * keys = (8,7,6), values = (1,2,2) --> keyList = (6,8), valueList = (2,1)</tt>
+ * </pre>
+ *
+ * @param condition the condition to be matched. Takes the current key as first and the current value as second
+ * argument.
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsMatching(final IntObjectProcedure condition, final IntArrayList keyList,
+ final ObjectArrayList valueList) {
+ keyList.clear();
+ valueList.clear();
+
+ forEachPair(
+ new IntObjectProcedure() {
+ @Override
+ public boolean apply(int key, Object value) {
+ if (condition.apply(key, value)) {
+ keyList.add(key);
+ valueList.add(value);
+ }
+ return true;
+ }
+ }
+ );
+ }
+
+ /**
+ * Fills all keys and values <i>sorted ascending by key</i> into the specified lists. Fills into the lists, starting
+ * at index 0. After this call returns the specified lists both have a new size that equals <tt>this.size()</tt>. <p>
+ * <b>Example:</b> <br> <tt>keys = (8,7,6), values = (1,2,2) --> keyList = (6,7,8), valueList = (2,2,1)</tt>
+ *
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsSortedByKey(IntArrayList keyList, ObjectArrayList valueList) {
+ keys(keyList);
+ keyList.sort();
+ valueList.setSize(keyList.size());
+ for (int i = keyList.size(); --i >= 0;) {
+ valueList.setQuick(i, get(keyList.getQuick(i)));
+ }
+ }
+
+ /**
+ * Fills all keys and values <i>sorted ascending by value according to natural ordering</i> into the specified lists.
+ * Fills into the lists, starting at index 0. After this call returns the specified lists both have a new size that
+ * equals <tt>this.size()</tt>. Primary sort criterium is "value", secondary sort criterium is "key". This means that
+ * if any two values are equal, the smaller key comes first. <p> <b>Example:</b> <br> <tt>keys = (8,7,6), values =
+ * (1,2,2) --> keyList = (8,6,7), valueList = (1,2,2)</tt>
+ *
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsSortedByValue(IntArrayList keyList, ObjectArrayList valueList) {
+ keys(keyList);
+ values(valueList);
+
+ final int[] k = keyList.elements();
+ final Object[] v = valueList.elements();
+ Swapper swapper = new Swapper() {
+ @Override
+ public void swap(int a, int b) {
+ Object t1 = v[a];
+ v[a] = v[b];
+ v[b] = t1;
+ int t2 = k[a];
+ k[a] = k[b];
+ k[b] = t2;
+ }
+ };
+
+ IntComparator comp = new IntComparator() {
+ @Override
+ public int compare(int a, int b) {
+ int ab = ((Comparable<Object>) v[a]).compareTo(v[b]);
+ return ab < 0 ? -1 : ab > 0 ? 1 : (k[a] < k[b] ? -1 : (k[a] == k[b] ? 0 : 1));
+ //return v[a]<v[b] ? -1 : v[a]>v[b] ? 1 : (k[a]<k[b] ? -1 : (k[a]==k[b] ? 0 : 1));
+ }
+ };
+
+ GenericSorting.quickSort(0, keyList.size(), comp, swapper);
+ }
+
+ /**
+ * Associates the given key with the given value. Replaces any old <tt>(key,someOtherValue)</tt> association, if
+ * existing.
+ *
+ * @param key the key the value shall be associated with.
+ * @param value the value to be associated.
+ * @return <tt>true</tt> if the receiver did not already contain such a key; <tt>false</tt> if the receiver did
+ * already contain such a key - the new value has now replaced the formerly associated value.
+ */
+ public abstract boolean put(int key, Object value);
+
+ /**
+ * Removes the given key with its associated element from the receiver, if present.
+ *
+ * @param key the key to be removed from the receiver.
+ * @return <tt>true</tt> if the receiver contained the specified key, <tt>false</tt> otherwise.
+ */
+ public abstract boolean removeKey(int key);
+
+ /**
+ * Returns a string representation of the receiver, containing the String representation of each key-value pair,
+ * sorted ascending by key.
+ */
+ public String toString() {
+ IntArrayList theKeys = keys();
+ theKeys.sort();
+
+ StringBuilder buf = new StringBuilder();
+ buf.append('[');
+ int maxIndex = theKeys.size() - 1;
+ for (int i = 0; i <= maxIndex; i++) {
+ int key = theKeys.get(i);
+ buf.append(String.valueOf(key));
+ buf.append("->");
+ buf.append(String.valueOf(get(key)));
+ if (i < maxIndex) {
+ buf.append(", ");
+ }
+ }
+ buf.append(']');
+ return buf.toString();
+ }
+
+ /**
+ * Returns a string representation of the receiver, containing the String representation of each key-value pair,
+ * sorted ascending by value, according to natural ordering.
+ */
+ public String toStringByValue() {
+ IntArrayList theKeys = new IntArrayList();
+ keysSortedByValue(theKeys);
+
+ StringBuilder buf = new StringBuilder();
+ buf.append('[');
+ int maxIndex = theKeys.size() - 1;
+ for (int i = 0; i <= maxIndex; i++) {
+ int key = theKeys.get(i);
+ buf.append(String.valueOf(key));
+ buf.append("->");
+ buf.append(String.valueOf(get(key)));
+ if (i < maxIndex) {
+ buf.append(", ");
+ }
+ }
+ buf.append(']');
+ return buf.toString();
+ }
+
+ /**
+ * Returns a list filled with all values contained in the receiver. The returned list has a size that equals
+ * <tt>this.size()</tt>. Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(IntProcedure)}. <p> This method can be used to iterate over the values of the receiver.
+ *
+ * @return the values.
+ */
+ public ObjectArrayList values() {
+ ObjectArrayList list = new ObjectArrayList(size());
+ values(list);
+ return list;
+ }
+
+ /**
+ * Fills all values contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(IntProcedure)}. <p> This method can be used to
+ * iterate over the values of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ public void values(final ObjectArrayList list) {
+ list.clear();
+ forEachKey(
+ new IntProcedure() {
+ @Override
+ public boolean apply(int key) {
+ list.add(get(key));
+ return true;
+ }
+ }
+ );
+ }
+}
Propchange: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractIntObjectMap.java
------------------------------------------------------------------------------
svn:eol-style = native
Added: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractLongObjectMap.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractLongObjectMap.java?rev=891983&view=auto
==============================================================================
--- lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractLongObjectMap.java (added)
+++ lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractLongObjectMap.java Thu Dec 17 23:22:16 2009
@@ -0,0 +1,442 @@
+/*
+Copyright � 1999 CERN - European Organization for Nuclear Research.
+Permission to use, copy, modify, distribute and sell this software and its documentation for any purpose
+is hereby granted without fee, provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear in supporting documentation.
+CERN makes no representations about the suitability of this software for any purpose.
+It is provided "as is" without expressed or implied warranty.
+*/
+package org.apache.mahout.math.map;
+
+import org.apache.mahout.math.GenericSorting;
+import org.apache.mahout.math.Swapper;
+import org.apache.mahout.math.function.IntComparator;
+import org.apache.mahout.math.function.LongObjectProcedure;
+import org.apache.mahout.math.function.LongProcedure;
+import org.apache.mahout.math.list.LongArrayList;
+import org.apache.mahout.math.list.ObjectArrayList;
+
+/** @deprecated until unit tests are in place. Until this time, this class/interface is unsupported. */
+@Deprecated
+public abstract class AbstractLongObjectMap extends AbstractMap {
+ //public static int hashCollisions = 0; // for debug only
+
+ /** Makes this class non instantiable, but still let's others inherit from it. */
+ protected AbstractLongObjectMap() {
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified key.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified key.
+ */
+ public boolean containsKey(final long key) {
+ return !forEachKey(
+ new LongProcedure() {
+ @Override
+ public boolean apply(long iterKey) {
+ return (key != iterKey);
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified value. Tests for identity.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified value.
+ */
+ public boolean containsValue(final Object value) {
+ return !forEachPair(
+ new LongObjectProcedure() {
+ @Override
+ public boolean apply(long iterKey, Object iterValue) {
+ return (value != iterValue);
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns a deep copy of the receiver; uses <code>clone()</code> and casts the result.
+ *
+ * @return a deep copy of the receiver.
+ */
+ public AbstractLongObjectMap copy() {
+ return (AbstractLongObjectMap) clone();
+ }
+
+ /**
+ * Compares the specified object with this map for equality. Returns <tt>true</tt> if the given object is also a map
+ * and the two maps represent the same mappings. More formally, two maps <tt>m1</tt> and <tt>m2</tt> represent the
+ * same mappings iff
+ * <pre>
+ * m1.forEachPair(
+ * new LongObjectProcedure() {
+ * public boolean apply(long key, Object value) {
+ * return m2.containsKey(key) && m2.get(key) == value;
+ * }
+ * }
+ * )
+ * &&
+ * m2.forEachPair(
+ * new LongObjectProcedure() {
+ * public boolean apply(long key, Object value) {
+ * return m1.containsKey(key) && m1.get(key) == value;
+ * }
+ * }
+ * );
+ * </pre>
+ *
+ * This implementation first checks if the specified object is this map; if so it returns <tt>true</tt>. Then, it
+ * checks if the specified object is a map whose size is identical to the size of this set; if not, it it returns
+ * <tt>false</tt>. If so, it applies the iteration as described above.
+ *
+ * @param obj object to be compared for equality with this map.
+ * @return <tt>true</tt> if the specified object is equal to this map.
+ */
+ public boolean equals(Object obj) {
+ if (obj == this) {
+ return true;
+ }
+
+ if (!(obj instanceof AbstractLongObjectMap)) {
+ return false;
+ }
+ final AbstractLongObjectMap other = (AbstractLongObjectMap) obj;
+ if (other.size() != size()) {
+ return false;
+ }
+
+ return
+ forEachPair(
+ new LongObjectProcedure() {
+ @Override
+ public boolean apply(long key, Object value) {
+ return other.containsKey(key) && other.get(key) == value;
+ }
+ }
+ )
+ &&
+ other.forEachPair(
+ new LongObjectProcedure() {
+ @Override
+ public boolean apply(long key, Object value) {
+ return containsKey(key) && get(key) == value;
+ }
+ }
+ );
+ }
+
+ /**
+ * Applies a procedure to each key of the receiver, if any. Note: Iterates over the keys in no particular order.
+ * Subclasses can define a particular order, for example, "sorted by key". All methods which <i>can</i> be expressed
+ * in terms of this method (most methods can) <i>must guarantee</i> to use the <i>same</i> order defined by this
+ * method, even if it is no particular order. This is necessary so that, for example, methods <tt>keys</tt> and
+ * <tt>values</tt> will yield association pairs, not two uncorrelated lists.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ public abstract boolean forEachKey(LongProcedure procedure);
+
+ /**
+ * Applies a procedure to each (key,value) pair of the receiver, if any. Iteration order is guaranteed to be
+ * <i>identical</i> to the order used by method {@link #forEachKey(LongProcedure)}.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ public boolean forEachPair(final LongObjectProcedure procedure) {
+ return forEachKey(
+ new LongProcedure() {
+ @Override
+ public boolean apply(long key) {
+ return procedure.apply(key, get(key));
+ }
+ }
+ );
+ }
+
+ /**
+ * Returns the value associated with the specified key. It is often a good idea to first check with {@link
+ * #containsKey(long)} whether the given key has a value associated or not, i.e. whether there exists an association
+ * for the given key or not.
+ *
+ * @param key the key to be searched for.
+ * @return the value associated with the specified key; <tt>null</tt> if no such key is present.
+ */
+ public abstract Object get(long key);
+
+ /**
+ * Returns the first key the given value is associated with. It is often a good idea to first check with {@link
+ * #containsValue(Object)} whether there exists an association from a key to this value. Search order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(LongProcedure)}.
+ *
+ * @param value the value to search for.
+ * @return the first key for which holds <tt>get(key) == value</tt>; returns <tt>Long.MIN_VALUE</tt> if no such key
+ * exists.
+ */
+ public long keyOf(final Object value) {
+ final long[] foundKey = new long[1];
+ boolean notFound = forEachPair(
+ new LongObjectProcedure() {
+ @Override
+ public boolean apply(long iterKey, Object iterValue) {
+ boolean found = value == iterValue;
+ if (found) {
+ foundKey[0] = iterKey;
+ }
+ return !found;
+ }
+ }
+ );
+ if (notFound) {
+ return Long.MIN_VALUE;
+ }
+ return foundKey[0];
+ }
+
+ /**
+ * Returns a list filled with all keys contained in the receiver. The returned list has a size that equals
+ * <tt>this.size()</tt>. Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(LongProcedure)}. <p> This method can be used to iterate over the keys of the receiver.
+ *
+ * @return the keys.
+ */
+ public LongArrayList keys() {
+ LongArrayList list = new LongArrayList(size());
+ keys(list);
+ return list;
+ }
+
+ /**
+ * Fills all keys contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(LongProcedure)}. <p> This method can be used to
+ * iterate over the keys of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ public void keys(final LongArrayList list) {
+ list.clear();
+ forEachKey(
+ new LongProcedure() {
+ @Override
+ public boolean apply(long key) {
+ list.add(key);
+ return true;
+ }
+ }
+ );
+ }
+
+ /**
+ * Fills all keys <i>sorted ascending by their associated value</i> into the specified list. Fills into the list,
+ * starting at index 0. After this call returns the specified list has a new size that equals <tt>this.size()</tt>.
+ * Primary sort criterium is "value", secondary sort criterium is "key". This means that if any two values are equal,
+ * the smaller key comes first. <p> <b>Example:</b> <br> <tt>keys = (8,7,6), values = (1,2,2) --> keyList =
+ * (8,6,7)</tt>
+ *
+ * @param keyList the list to be filled, can have any size.
+ */
+ public void keysSortedByValue(LongArrayList keyList) {
+ pairsSortedByValue(keyList, new ObjectArrayList(size()));
+ }
+
+ /**
+ * Fills all pairs satisfying a given condition into the specified lists. Fills into the lists, starting at index 0.
+ * After this call returns the specified lists both have a new size, the number of pairs satisfying the condition.
+ * Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(LongProcedure)}. <p> <b>Example:</b> <br>
+ * <pre>
+ * LongObjectProcedure condition = new LongObjectProcedure() { // match even keys only
+ * public boolean apply(long key, Object value) { return key%2==0; }
+ * }
+ * keys = (8,7,6), values = (1,2,2) --> keyList = (6,8), valueList = (2,1)</tt>
+ * </pre>
+ *
+ * @param condition the condition to be matched. Takes the current key as first and the current value as second
+ * argument.
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsMatching(final LongObjectProcedure condition, final LongArrayList keyList,
+ final ObjectArrayList valueList) {
+ keyList.clear();
+ valueList.clear();
+
+ forEachPair(
+ new LongObjectProcedure() {
+ @Override
+ public boolean apply(long key, Object value) {
+ if (condition.apply(key, value)) {
+ keyList.add(key);
+ valueList.add(value);
+ }
+ return true;
+ }
+ }
+ );
+ }
+
+ /**
+ * Fills all keys and values <i>sorted ascending by key</i> into the specified lists. Fills into the lists, starting
+ * at index 0. After this call returns the specified lists both have a new size that equals <tt>this.size()</tt>. <p>
+ * <b>Example:</b> <br> <tt>keys = (8,7,6), values = (1,2,2) --> keyList = (6,7,8), valueList = (2,2,1)</tt>
+ *
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsSortedByKey(LongArrayList keyList, ObjectArrayList valueList) {
+ keys(keyList);
+ keyList.sort();
+ valueList.setSize(keyList.size());
+ for (int i = keyList.size(); --i >= 0;) {
+ valueList.setQuick(i, get(keyList.getQuick(i)));
+ }
+ }
+
+ /**
+ * Fills all keys and values <i>sorted ascending by value according to natural ordering</i> into the specified lists.
+ * Fills into the lists, starting at index 0. After this call returns the specified lists both have a new size that
+ * equals <tt>this.size()</tt>. Primary sort criterium is "value", secondary sort criterium is "key". This means that
+ * if any two values are equal, the smaller key comes first. <p> <b>Example:</b> <br> <tt>keys = (8,7,6), values =
+ * (1,2,2) --> keyList = (8,6,7), valueList = (1,2,2)</tt>
+ *
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ public void pairsSortedByValue(LongArrayList keyList, ObjectArrayList valueList) {
+ keys(keyList);
+ values(valueList);
+
+ final long[] k = keyList.elements();
+ final Object[] v = valueList.elements();
+ Swapper swapper = new Swapper() {
+ @Override
+ public void swap(int a, int b) {
+ Object t1 = v[a];
+ v[a] = v[b];
+ v[b] = t1;
+ long t2 = k[a];
+ k[a] = k[b];
+ k[b] = t2;
+ }
+ };
+
+ IntComparator comp = new IntComparator() {
+ @Override
+ public int compare(int a, int b) {
+ int ab = ((Comparable<Object>) v[a]).compareTo(v[b]);
+ return ab < 0 ? -1 : ab > 0 ? 1 : (k[a] < k[b] ? -1 : (k[a] == k[b] ? 0 : 1));
+ //return v[a]<v[b] ? -1 : v[a]>v[b] ? 1 : (k[a]<k[b] ? -1 : (k[a]==k[b] ? 0 : 1));
+ }
+ };
+
+ GenericSorting.quickSort(0, keyList.size(), comp, swapper);
+ }
+
+ /**
+ * Associates the given key with the given value. Replaces any old <tt>(key,someOtherValue)</tt> association, if
+ * existing.
+ *
+ * @param key the key the value shall be associated with.
+ * @param value the value to be associated.
+ * @return <tt>true</tt> if the receiver did not already contain such a key; <tt>false</tt> if the receiver did
+ * already contain such a key - the new value has now replaced the formerly associated value.
+ */
+ public abstract boolean put(long key, Object value);
+
+ /**
+ * Removes the given key with its associated element from the receiver, if present.
+ *
+ * @param key the key to be removed from the receiver.
+ * @return <tt>true</tt> if the receiver contained the specified key, <tt>false</tt> otherwise.
+ */
+ public abstract boolean removeKey(long key);
+
+ /**
+ * Returns a string representation of the receiver, containing the String representation of each key-value pair,
+ * sorted ascending by key.
+ */
+ public String toString() {
+ LongArrayList theKeys = keys();
+ theKeys.sort();
+
+ StringBuilder buf = new StringBuilder();
+ buf.append('[');
+ int maxIndex = theKeys.size() - 1;
+ for (int i = 0; i <= maxIndex; i++) {
+ long key = theKeys.get(i);
+ buf.append(String.valueOf(key));
+ buf.append("->");
+ buf.append(String.valueOf(get(key)));
+ if (i < maxIndex) {
+ buf.append(", ");
+ }
+ }
+ buf.append(']');
+ return buf.toString();
+ }
+
+ /**
+ * Returns a string representation of the receiver, containing the String representation of each key-value pair,
+ * sorted ascending by value, according to natural ordering.
+ */
+ public String toStringByValue() {
+ LongArrayList theKeys = new LongArrayList();
+ keysSortedByValue(theKeys);
+
+ StringBuilder buf = new StringBuilder();
+ buf.append('[');
+ int maxIndex = theKeys.size() - 1;
+ for (int i = 0; i <= maxIndex; i++) {
+ long key = theKeys.get(i);
+ buf.append(String.valueOf(key));
+ buf.append("->");
+ buf.append(String.valueOf(get(key)));
+ if (i < maxIndex) {
+ buf.append(", ");
+ }
+ }
+ buf.append(']');
+ return buf.toString();
+ }
+
+ /**
+ * Returns a list filled with all values contained in the receiver. The returned list has a size that equals
+ * <tt>this.size()</tt>. Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(LongProcedure)}. <p> This method can be used to iterate over the values of the receiver.
+ *
+ * @return the values.
+ */
+ public ObjectArrayList values() {
+ ObjectArrayList list = new ObjectArrayList(size());
+ values(list);
+ return list;
+ }
+
+ /**
+ * Fills all values contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(LongProcedure)}. <p> This method can be used to
+ * iterate over the values of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ public void values(final ObjectArrayList list) {
+ list.clear();
+ forEachKey(
+ new LongProcedure() {
+ @Override
+ public boolean apply(long key) {
+ list.add(get(key));
+ return true;
+ }
+ }
+ );
+ }
+}
Propchange: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractLongObjectMap.java
------------------------------------------------------------------------------
svn:eol-style = native
Added: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractMap.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractMap.java?rev=891983&view=auto
==============================================================================
--- lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractMap.java (added)
+++ lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractMap.java Thu Dec 17 23:22:16 2009
@@ -0,0 +1,187 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+/*
+Copyright 1999 CERN - European Organization for Nuclear Research.
+Permission to use, copy, modify, distribute and sell this software and its documentation for any purpose
+is hereby granted without fee, provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear in supporting documentation.
+CERN makes no representations about the suitability of this software for any purpose.
+It is provided "as is" without expressed or implied warranty.
+*/
+package org.apache.mahout.math.map;
+
+import org.apache.mahout.math.PersistentObject;
+
+/**
+ Abstract base class for hash maps holding objects or primitive data types such as <code>int</code>, <code>float</code>, etc. as keys and/or values.
+ First see the <a href="package-summary.html">package summary</a>
+ and javadoc <a href="package-tree.html">tree view</a> to get the broad picture.
+ <p>
+ Note that implementations are not synchronized.
+
+ @author wolfgang.hoschek@cern.ch
+ @version 1.0, 09/24/99
+ @see java.util.HashMap
+ */
+public abstract class AbstractMap extends PersistentObject {
+ //public static boolean debug = false; // debug only
+
+ /** The number of distinct associations in the map; its "size()". */
+ protected int distinct;
+
+ /**
+ * The table capacity c=table.length always satisfies the invariant <tt>c * minLoadFactor <= s <= c *
+ * maxLoadFactor</tt>, where s=size() is the number of associations currently contained. The term "c * minLoadFactor"
+ * is called the "lowWaterMark", "c * maxLoadFactor" is called the "highWaterMark". In other words, the table capacity
+ * (and proportionally the memory used by this class) oscillates within these constraints. The terms are precomputed
+ * and cached to avoid recalculating them each time put(..) or removeKey(...) is called.
+ */
+ protected int lowWaterMark;
+ protected int highWaterMark;
+
+ /** The minimum load factor for the hashtable. */
+ protected double minLoadFactor;
+
+ /** The maximum load factor for the hashtable. */
+ protected double maxLoadFactor;
+
+ protected static final int defaultCapacity = 277;
+ protected static final double defaultMinLoadFactor = 0.2;
+ protected static final double defaultMaxLoadFactor = 0.5;
+
+ /**
+ * Chooses a new prime table capacity optimized for growing that (approximately) satisfies the invariant <tt>c *
+ * minLoadFactor <= size <= c * maxLoadFactor</tt> and has at least one FREE slot for the given size.
+ */
+ protected int chooseGrowCapacity(int size, double minLoad, double maxLoad) {
+ return nextPrime(Math.max(size + 1, (int) ((4 * size / (3 * minLoad + maxLoad)))));
+ }
+
+ /**
+ * Returns new high water mark threshold based on current capacity and maxLoadFactor.
+ *
+ * @return int the new threshold.
+ */
+ protected int chooseHighWaterMark(int capacity, double maxLoad) {
+ return Math.min(capacity - 2, (int) (capacity * maxLoad)); //makes sure there is always at least one FREE slot
+ }
+
+ /**
+ * Returns new low water mark threshold based on current capacity and minLoadFactor.
+ *
+ * @return int the new threshold.
+ */
+ protected int chooseLowWaterMark(int capacity, double minLoad) {
+ return (int) (capacity * minLoad);
+ }
+
+ /**
+ * Chooses a new prime table capacity neither favoring shrinking nor growing, that (approximately) satisfies the
+ * invariant <tt>c * minLoadFactor <= size <= c * maxLoadFactor</tt> and has at least one FREE slot for the given
+ * size.
+ */
+ protected int chooseMeanCapacity(int size, double minLoad, double maxLoad) {
+ return nextPrime(Math.max(size + 1, (int) ((2 * size / (minLoad + maxLoad)))));
+ }
+
+ /**
+ * Chooses a new prime table capacity optimized for shrinking that (approximately) satisfies the invariant <tt>c *
+ * minLoadFactor <= size <= c * maxLoadFactor</tt> and has at least one FREE slot for the given size.
+ */
+ protected int chooseShrinkCapacity(int size, double minLoad, double maxLoad) {
+ return nextPrime(Math.max(size + 1, (int) ((4 * size / (minLoad + 3 * maxLoad)))));
+ }
+
+ /** Removes all (key,value) associations from the receiver. */
+ public abstract void clear();
+
+ /**
+ * Ensures that the receiver can hold at least the specified number of elements without needing to allocate new
+ * internal memory. If necessary, allocates new internal memory and increases the capacity of the receiver. <p> This
+ * method never need be called; it is for performance tuning only. Calling this method before <tt>put()</tt>ing a
+ * large number of associations boosts performance, because the receiver will grow only once instead of potentially
+ * many times. <p> <b>This default implementation does nothing.</b> Override this method if necessary.
+ *
+ * @param minCapacity the desired minimum capacity.
+ */
+ public void ensureCapacity(int minCapacity) {
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains no (key,value) associations.
+ *
+ * @return <tt>true</tt> if the receiver contains no (key,value) associations.
+ */
+ public boolean isEmpty() {
+ return distinct == 0;
+ }
+
+ /**
+ * Returns a prime number which is <code>>= desiredCapacity</code> and very close to <code>desiredCapacity</code>
+ * (within 11% if <code>desiredCapacity >= 1000</code>).
+ *
+ * @param desiredCapacity the capacity desired by the user.
+ * @return the capacity which should be used for a hashtable.
+ */
+protected int nextPrime(int desiredCapacity) {
+ return PrimeFinder.nextPrime(desiredCapacity);
+ }
+
+ /**
+ * Initializes the receiver. You will almost certainly need to override this method in subclasses to initialize the
+ * hash table.
+ *
+ * @param initialCapacity the initial capacity of the receiver.
+ * @param minLoadFactor the minLoadFactor of the receiver.
+ * @param maxLoadFactor the maxLoadFactor of the receiver.
+ * @throws IllegalArgumentException if <tt>initialCapacity < 0 || (minLoadFactor < 0.0 || minLoadFactor >= 1.0) ||
+ * (maxLoadFactor <= 0.0 || maxLoadFactor >= 1.0) || (minLoadFactor >=
+ * maxLoadFactor)</tt>.
+ */
+ protected void setUp(int initialCapacity, double minLoadFactor, double maxLoadFactor) {
+ if (initialCapacity < 0) {
+ throw new IllegalArgumentException("Initial Capacity must not be less than zero: " + initialCapacity);
+ }
+ if (minLoadFactor < 0.0 || minLoadFactor >= 1.0) {
+ throw new IllegalArgumentException("Illegal minLoadFactor: " + minLoadFactor);
+ }
+ if (maxLoadFactor <= 0.0 || maxLoadFactor >= 1.0) {
+ throw new IllegalArgumentException("Illegal maxLoadFactor: " + maxLoadFactor);
+ }
+ if (minLoadFactor >= maxLoadFactor) {
+ throw new IllegalArgumentException(
+ "Illegal minLoadFactor: " + minLoadFactor + " and maxLoadFactor: " + maxLoadFactor);
+ }
+ }
+
+ /**
+ * Returns the number of (key,value) associations currently contained.
+ *
+ * @return the number of (key,value) associations currently contained.
+ */
+ public int size() {
+ return distinct;
+ }
+
+ /**
+ * Trims the capacity of the receiver to be the receiver's current size. Releases any superfluous internal memory. An
+ * application can use this operation to minimize the storage of the receiver. <p> This default implementation does
+ * nothing. Override this method if necessary.
+ */
+ public void trimToSize() {
+ }
+}
Propchange: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/AbstractMap.java
------------------------------------------------------------------------------
svn:eol-style = native
Added: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/HashFunctions.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/HashFunctions.java?rev=891983&view=auto
==============================================================================
--- lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/HashFunctions.java (added)
+++ lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/HashFunctions.java Thu Dec 17 23:22:16 2009
@@ -0,0 +1,119 @@
+/*
+Copyright 1999 CERN - European Organization for Nuclear Research.
+Permission to use, copy, modify, distribute and sell this software and its documentation for any purpose
+is hereby granted without fee, provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear in supporting documentation.
+CERN makes no representations about the suitability of this software for any purpose.
+It is provided "as is" without expressed or implied warranty.
+*/
+package org.apache.mahout.math.map;
+
+/**
+ * Provides various hash functions.
+ *
+ */
+
+/** @deprecated until unit tests are in place. Until this time, this class/interface is unsupported. */
+@Deprecated
+public class HashFunctions {
+
+ /** Makes this class non instantiable, but still let's others inherit from it. */
+ private HashFunctions() {
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(char value) {
+ return (int) value;
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(double value) {
+ long bits = Double.doubleToLongBits(value);
+ return (int) (bits ^ (bits >>> 32));
+
+ //return (int) Double.doubleToLongBits(value*663608941.737);
+ // this avoids excessive hashCollisions in the case values are of the form (1.0, 2.0, 3.0, ...)
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(float value) {
+ return Float.floatToIntBits(value * 663608941.737f);
+ // this avoids excessive hashCollisions in the case values are of the form (1.0, 2.0, 3.0, ...)
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(int value) {
+ return value;
+
+ //return value * 0x278DDE6D; // see org.apache.mahout.math.jet.random.engine.DRand
+
+ /*
+ value &= 0x7FFFFFFF; // make it >=0
+ int hashCode = 0;
+ do hashCode = 31*hashCode + value%10;
+ while ((value /= 10) > 0);
+
+ return 28629151*hashCode; // spread even further; h*31^5
+ */
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(long value) {
+ return (int) (value ^ (value >> 32));
+ /*
+ value &= 0x7FFFFFFFFFFFFFFFL; // make it >=0 (0x7FFFFFFFFFFFFFFFL==Long.MAX_VALUE)
+ int hashCode = 0;
+ do hashCode = 31*hashCode + (int) (value%10);
+ while ((value /= 10) > 0);
+
+ return 28629151*hashCode; // spread even further; h*31^5
+ */
+ }
+
+ /**
+ * Returns a hashcode for the specified object.
+ *
+ * @return a hash code value for the specified object.
+ */
+ public static int hash(Object object) {
+ return object == null ? 0 : object.hashCode();
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(short value) {
+ return (int) value;
+ }
+
+ /**
+ * Returns a hashcode for the specified value.
+ *
+ * @return a hash code value for the specified value.
+ */
+ public static int hash(boolean value) {
+ return value ? 1231 : 1237;
+ }
+}
Propchange: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/HashFunctions.java
------------------------------------------------------------------------------
svn:eol-style = native
Added: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/OpenDoubleIntHashMap.java
URL: http://svn.apache.org/viewvc/lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/OpenDoubleIntHashMap.java?rev=891983&view=auto
==============================================================================
--- lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/OpenDoubleIntHashMap.java (added)
+++ lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/OpenDoubleIntHashMap.java Thu Dec 17 23:22:16 2009
@@ -0,0 +1,560 @@
+/*
+Copyright � 1999 CERN - European Organization for Nuclear Research.
+Permission to use, copy, modify, distribute and sell this software and its documentation for any purpose
+is hereby granted without fee, provided that the above copyright notice appear in all copies and
+that both that copyright notice and this permission notice appear in supporting documentation.
+CERN makes no representations about the suitability of this software for any purpose.
+It is provided "as is" without expressed or implied warranty.
+*/
+package org.apache.mahout.math.map;
+
+import org.apache.mahout.math.function.DoubleIntProcedure;
+import org.apache.mahout.math.function.DoubleProcedure;
+import org.apache.mahout.math.list.ByteArrayList;
+import org.apache.mahout.math.list.DoubleArrayList;
+import org.apache.mahout.math.list.IntArrayList;
+
+/** @deprecated until unit tests are in place. Until this time, this class/interface is unsupported. */
+@Deprecated
+public class OpenDoubleIntHashMap extends AbstractDoubleIntMap {
+
+ /** The hash table keys. */
+ private double[] table;
+
+ /** The hash table values. */
+ private int[] values;
+
+ /** The state of each hash table entry (FREE, FULL, REMOVED). */
+ private byte[] state;
+
+ /** The number of table entries in state==FREE. */
+ private int freeEntries;
+
+
+ private static final byte FREE = 0;
+ private static final byte FULL = 1;
+ private static final byte REMOVED = 2;
+
+ /** Constructs an empty map with default capacity and default load factors. */
+ public OpenDoubleIntHashMap() {
+ this(defaultCapacity);
+ }
+
+ /**
+ * Constructs an empty map with the specified initial capacity and default load factors.
+ *
+ * @param initialCapacity the initial capacity of the map.
+ * @throws IllegalArgumentException if the initial capacity is less than zero.
+ */
+ public OpenDoubleIntHashMap(int initialCapacity) {
+ this(initialCapacity, defaultMinLoadFactor, defaultMaxLoadFactor);
+ }
+
+ /**
+ * Constructs an empty map with the specified initial capacity and the specified minimum and maximum load factor.
+ *
+ * @param initialCapacity the initial capacity.
+ * @param minLoadFactor the minimum load factor.
+ * @param maxLoadFactor the maximum load factor.
+ * @throws IllegalArgumentException if <tt>initialCapacity < 0 || (minLoadFactor < 0.0 || minLoadFactor >= 1.0) ||
+ * (maxLoadFactor <= 0.0 || maxLoadFactor >= 1.0) || (minLoadFactor >=
+ * maxLoadFactor)</tt>.
+ */
+ public OpenDoubleIntHashMap(int initialCapacity, double minLoadFactor, double maxLoadFactor) {
+ setUp(initialCapacity, minLoadFactor, maxLoadFactor);
+ }
+
+ /** Removes all (key,value) associations from the receiver. Implicitly calls <tt>trimToSize()</tt>. */
+ @Override
+ public void clear() {
+ new ByteArrayList(this.state).fillFromToWith(0, this.state.length - 1, FREE);
+ //new DoubleArrayList(values).fillFromToWith(0, state.length-1, 0); // delta
+
+ this.distinct = 0;
+ this.freeEntries = table.length; // delta
+ trimToSize();
+ }
+
+ /**
+ * Returns a deep copy of the receiver.
+ *
+ * @return a deep copy of the receiver.
+ */
+ @Override
+ public Object clone() {
+ OpenDoubleIntHashMap copy = (OpenDoubleIntHashMap) super.clone();
+ copy.table = copy.table.clone();
+ copy.values = copy.values.clone();
+ copy.state = copy.state.clone();
+ return copy;
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified key.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified key.
+ */
+ @Override
+ public boolean containsKey(double key) {
+ return indexOfKey(key) >= 0;
+ }
+
+ /**
+ * Returns <tt>true</tt> if the receiver contains the specified value.
+ *
+ * @return <tt>true</tt> if the receiver contains the specified value.
+ */
+ @Override
+ public boolean containsValue(int value) {
+ return indexOfValue(value) >= 0;
+ }
+
+ /**
+ * Ensures that the receiver can hold at least the specified number of associations without needing to allocate new
+ * internal memory. If necessary, allocates new internal memory and increases the capacity of the receiver. <p> This
+ * method never need be called; it is for performance tuning only. Calling this method before <tt>put()</tt>ing a
+ * large number of associations boosts performance, because the receiver will grow only once instead of potentially
+ * many times and hash collisions get less probable.
+ *
+ * @param minCapacity the desired minimum capacity.
+ */
+ @Override
+ public void ensureCapacity(int minCapacity) {
+ if (table.length < minCapacity) {
+ int newCapacity = nextPrime(minCapacity);
+ rehash(newCapacity);
+ }
+ }
+
+ /**
+ * Applies a procedure to each key of the receiver, if any. Note: Iterates over the keys in no particular order.
+ * Subclasses can define a particular order, for example, "sorted by key". All methods which <i>can</i> be expressed
+ * in terms of this method (most methods can) <i>must guarantee</i> to use the <i>same</i> order defined by this
+ * method, even if it is no particular order. This is necessary so that, for example, methods <tt>keys</tt> and
+ * <tt>values</tt> will yield association pairs, not two uncorrelated lists.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ @Override
+ public boolean forEachKey(DoubleProcedure procedure) {
+ for (int i = table.length; i-- > 0;) {
+ if (state[i] == FULL) {
+ if (!procedure.apply(table[i])) {
+ return false;
+ }
+ }
+ }
+ return true;
+ }
+
+ /**
+ * Applies a procedure to each (key,value) pair of the receiver, if any. Iteration order is guaranteed to be
+ * <i>identical</i> to the order used by method {@link #forEachKey(DoubleProcedure)}.
+ *
+ * @param procedure the procedure to be applied. Stops iteration if the procedure returns <tt>false</tt>, otherwise
+ * continues.
+ * @return <tt>false</tt> if the procedure stopped before all keys where iterated over, <tt>true</tt> otherwise.
+ */
+ @Override
+ public boolean forEachPair(DoubleIntProcedure procedure) {
+ for (int i = table.length; i-- > 0;) {
+ if (state[i] == FULL) {
+ if (!procedure.apply(table[i], values[i])) {
+ return false;
+ }
+ }
+ }
+ return true;
+ }
+
+ /**
+ * Returns the value associated with the specified key. It is often a good idea to first check with {@link
+ * #containsKey(double)} whether the given key has a value associated or not, i.e. whether there exists an association
+ * for the given key or not.
+ *
+ * @param key the key to be searched for.
+ * @return the value associated with the specified key; <tt>0</tt> if no such key is present.
+ */
+ @Override
+ public int get(double key) {
+ int i = indexOfKey(key);
+ if (i < 0) {
+ return 0;
+ } //not contained
+ return values[i];
+ }
+
+ /**
+ * @param key the key to be added to the receiver.
+ * @return the index where the key would need to be inserted, if it is not already contained. Returns -index-1 if the
+ * key is already contained at slot index. Therefore, if the returned index < 0, then it is already contained
+ * at slot -index-1. If the returned index >= 0, then it is NOT already contained and should be inserted at
+ * slot index.
+ */
+ protected int indexOfInsertion(double key) {
+ double[] tab = table;
+ byte[] stat = state;
+ int length = tab.length;
+
+ int hash = HashFunctions.hash(key) & 0x7FFFFFFF;
+ int i = hash % length;
+ int decrement = hash % (length - 2); // double hashing, see http://www.eece.unm.edu/faculty/heileman/hash/node4.html
+ //int decrement = (hash / length) % length;
+ if (decrement == 0) {
+ decrement = 1;
+ }
+
+ // stop if we find a removed or free slot, or if we find the key itself
+ // do NOT skip over removed slots (yes, open addressing is like that...)
+ while (stat[i] == FULL && tab[i] != key) {
+ i -= decrement;
+ //hashCollisions++;
+ if (i < 0) {
+ i += length;
+ }
+ }
+
+ if (stat[i] == REMOVED) {
+ // stop if we find a free slot, or if we find the key itself.
+ // do skip over removed slots (yes, open addressing is like that...)
+ // assertion: there is at least one FREE slot.
+ int j = i;
+ while (stat[i] != FREE && (stat[i] == REMOVED || tab[i] != key)) {
+ i -= decrement;
+ //hashCollisions++;
+ if (i < 0) {
+ i += length;
+ }
+ }
+ if (stat[i] == FREE) {
+ i = j;
+ }
+ }
+
+
+ if (stat[i] == FULL) {
+ // key already contained at slot i.
+ // return a negative number identifying the slot.
+ return -i - 1;
+ }
+ // not already contained, should be inserted at slot i.
+ // return a number >= 0 identifying the slot.
+ return i;
+ }
+
+ /**
+ * @param key the key to be searched in the receiver.
+ * @return the index where the key is contained in the receiver, returns -1 if the key was not found.
+ */
+ protected int indexOfKey(double key) {
+ double[] tab = table;
+ byte[] stat = state;
+ int length = tab.length;
+
+ int hash = HashFunctions.hash(key) & 0x7FFFFFFF;
+ int i = hash % length;
+ int decrement = hash % (length - 2); // double hashing, see http://www.eece.unm.edu/faculty/heileman/hash/node4.html
+ //int decrement = (hash / length) % length;
+ if (decrement == 0) {
+ decrement = 1;
+ }
+
+ // stop if we find a free slot, or if we find the key itself.
+ // do skip over removed slots (yes, open addressing is like that...)
+ while (stat[i] != FREE && (stat[i] == REMOVED || tab[i] != key)) {
+ i -= decrement;
+ //hashCollisions++;
+ if (i < 0) {
+ i += length;
+ }
+ }
+
+ if (stat[i] == FREE) {
+ return -1;
+ } // not found
+ return i; //found, return index where key is contained
+ }
+
+ /**
+ * @param value the value to be searched in the receiver.
+ * @return the index where the value is contained in the receiver, returns -1 if the value was not found.
+ */
+ protected int indexOfValue(int value) {
+ int[] val = values;
+ byte[] stat = state;
+
+ for (int i = stat.length; --i >= 0;) {
+ if (stat[i] == FULL && val[i] == value) {
+ return i;
+ }
+ }
+
+ return -1; // not found
+ }
+
+ /**
+ * Returns the first key the given value is associated with. It is often a good idea to first check with {@link
+ * #containsValue(int)} whether there exists an association from a key to this value. Search order is guaranteed to be
+ * <i>identical</i> to the order used by method {@link #forEachKey(DoubleProcedure)}.
+ *
+ * @param value the value to search for.
+ * @return the first key for which holds <tt>get(key) == value</tt>; returns <tt>Double.NaN</tt> if no such key
+ * exists.
+ */
+ @Override
+ public double keyOf(int value) {
+ //returns the first key found; there may be more matching keys, however.
+ int i = indexOfValue(value);
+ if (i < 0) {
+ return Double.NaN;
+ }
+ return table[i];
+ }
+
+ /**
+ * Fills all keys contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(DoubleProcedure)}. <p> This method can be used
+ * to iterate over the keys of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ @Override
+ public void keys(DoubleArrayList list) {
+ list.setSize(distinct);
+ double[] elements = list.elements();
+
+ double[] tab = table;
+ byte[] stat = state;
+
+ int j = 0;
+ for (int i = tab.length; i-- > 0;) {
+ if (stat[i] == FULL) {
+ elements[j++] = tab[i];
+ }
+ }
+ }
+
+ /**
+ * Fills all pairs satisfying a given condition into the specified lists. Fills into the lists, starting at index 0.
+ * After this call returns the specified lists both have a new size, the number of pairs satisfying the condition.
+ * Iteration order is guaranteed to be <i>identical</i> to the order used by method {@link
+ * #forEachKey(DoubleProcedure)}. <p> <b>Example:</b> <br>
+ * <pre>
+ * DoubleIntProcedure condition = new DoubleIntProcedure() { // match even values only
+ * public boolean apply(double key, int value) { return value%2==0; }
+ * }
+ * keys = (8,7,6), values = (1,2,2) --> keyList = (6,8), valueList = (2,1)</tt>
+ * </pre>
+ *
+ * @param condition the condition to be matched. Takes the current key as first and the current value as second
+ * argument.
+ * @param keyList the list to be filled with keys, can have any size.
+ * @param valueList the list to be filled with values, can have any size.
+ */
+ @Override
+ public void pairsMatching(DoubleIntProcedure condition, DoubleArrayList keyList, IntArrayList valueList) {
+ keyList.clear();
+ valueList.clear();
+
+ for (int i = table.length; i-- > 0;) {
+ if (state[i] == FULL && condition.apply(table[i], values[i])) {
+ keyList.add(table[i]);
+ valueList.add(values[i]);
+ }
+ }
+ }
+
+ /**
+ * Associates the given key with the given value. Replaces any old <tt>(key,someOtherValue)</tt> association, if
+ * existing.
+ *
+ * @param key the key the value shall be associated with.
+ * @param value the value to be associated.
+ * @return <tt>true</tt> if the receiver did not already contain such a key; <tt>false</tt> if the receiver did
+ * already contain such a key - the new value has now replaced the formerly associated value.
+ */
+ @Override
+ public boolean put(double key, int value) {
+ int i = indexOfInsertion(key);
+ if (i < 0) { //already contained
+ i = -i - 1;
+ this.values[i] = value;
+ return false;
+ }
+
+ if (this.distinct > this.highWaterMark) {
+ int newCapacity = chooseGrowCapacity(this.distinct + 1, this.minLoadFactor, this.maxLoadFactor);
+ /*
+ log.info("grow rehashing ");
+ log.info("at distinct="+distinct+", capacity="+table.length+" to newCapacity="+newCapacity+" ...");
+ */
+ rehash(newCapacity);
+ return put(key, value);
+ }
+
+ this.table[i] = key;
+ this.values[i] = value;
+ if (this.state[i] == FREE) {
+ this.freeEntries--;
+ }
+ this.state[i] = FULL;
+ this.distinct++;
+
+ if (this.freeEntries < 1) { //delta
+ int newCapacity = chooseGrowCapacity(this.distinct + 1, this.minLoadFactor, this.maxLoadFactor);
+ rehash(newCapacity);
+ }
+
+ return true;
+ }
+
+ /**
+ * Rehashes the contents of the receiver into a new table with a smaller or larger capacity. This method is called
+ * automatically when the number of keys in the receiver exceeds the high water mark or falls below the low water
+ * mark.
+ */
+ protected void rehash(int newCapacity) {
+ int oldCapacity = table.length;
+ //if (oldCapacity == newCapacity) return;
+
+ double[] oldTable = table;
+ int[] oldValues = values;
+ byte[] oldState = state;
+
+ double[] newTable = new double[newCapacity];
+ int[] newValues = new int[newCapacity];
+ byte[] newState = new byte[newCapacity];
+
+ this.lowWaterMark = chooseLowWaterMark(newCapacity, this.minLoadFactor);
+ this.highWaterMark = chooseHighWaterMark(newCapacity, this.maxLoadFactor);
+
+ this.table = newTable;
+ this.values = newValues;
+ this.state = newState;
+ this.freeEntries = newCapacity - this.distinct; // delta
+
+ for (int i = oldCapacity; i-- > 0;) {
+ if (oldState[i] == FULL) {
+ double element = oldTable[i];
+ int index = indexOfInsertion(element);
+ newTable[index] = element;
+ newValues[index] = oldValues[i];
+ newState[index] = FULL;
+ }
+ }
+ }
+
+ /**
+ * Removes the given key with its associated element from the receiver, if present.
+ *
+ * @param key the key to be removed from the receiver.
+ * @return <tt>true</tt> if the receiver contained the specified key, <tt>false</tt> otherwise.
+ */
+ @Override
+ public boolean removeKey(double key) {
+ int i = indexOfKey(key);
+ if (i < 0) {
+ return false;
+ } // key not contained
+
+ this.state[i] = REMOVED;
+ //this.values[i]=0; // delta
+ this.distinct--;
+
+ if (this.distinct < this.lowWaterMark) {
+ int newCapacity = chooseShrinkCapacity(this.distinct, this.minLoadFactor, this.maxLoadFactor);
+ /*
+ if (table.length != newCapacity) {
+ log.info("shrink rehashing ");
+ log.info("at distinct="+distinct+", capacity="+table.length+" to newCapacity="+newCapacity+" ...");
+ }
+ */
+ rehash(newCapacity);
+ }
+
+ return true;
+ }
+
+ /**
+ * Initializes the receiver.
+ *
+ * @param initialCapacity the initial capacity of the receiver.
+ * @param minLoadFactor the minLoadFactor of the receiver.
+ * @param maxLoadFactor the maxLoadFactor of the receiver.
+ * @throws IllegalArgumentException if <tt>initialCapacity < 0 || (minLoadFactor < 0.0 || minLoadFactor >= 1.0) ||
+ * (maxLoadFactor <= 0.0 || maxLoadFactor >= 1.0) || (minLoadFactor >=
+ * maxLoadFactor)</tt>.
+ */
+ @Override
+ protected void setUp(int initialCapacity, double minLoadFactor, double maxLoadFactor) {
+ int capacity = initialCapacity;
+ super.setUp(capacity, minLoadFactor, maxLoadFactor);
+ capacity = nextPrime(capacity);
+ if (capacity == 0) {
+ capacity = 1;
+ } // open addressing needs at least one FREE slot at any time.
+
+ this.table = new double[capacity];
+ this.values = new int[capacity];
+ this.state = new byte[capacity];
+
+ // memory will be exhausted long before this pathological case happens, anyway.
+ this.minLoadFactor = minLoadFactor;
+ if (capacity == PrimeFinder.largestPrime) {
+ this.maxLoadFactor = 1.0;
+ } else {
+ this.maxLoadFactor = maxLoadFactor;
+ }
+
+ this.distinct = 0;
+ this.freeEntries = capacity; // delta
+
+ // lowWaterMark will be established upon first expansion.
+ // establishing it now (upon instance construction) would immediately make the table shrink upon first put(...).
+ // After all the idea of an "initialCapacity" implies violating lowWaterMarks when an object is young.
+ // See ensureCapacity(...)
+ this.lowWaterMark = 0;
+ this.highWaterMark = chooseHighWaterMark(capacity, this.maxLoadFactor);
+ }
+
+ /**
+ * Trims the capacity of the receiver to be the receiver's current size. Releases any superfluous internal memory. An
+ * application can use this operation to minimize the storage of the receiver.
+ */
+ @Override
+ public void trimToSize() {
+ // * 1.2 because open addressing's performance exponentially degrades beyond that point
+ // so that even rehashing the table can take very long
+ int newCapacity = nextPrime((int) (1 + 1.2 * size()));
+ if (table.length > newCapacity) {
+ rehash(newCapacity);
+ }
+ }
+
+ /**
+ * Fills all values contained in the receiver into the specified list. Fills the list, starting at index 0. After this
+ * call returns the specified list has a new size that equals <tt>this.size()</tt>. Iteration order is guaranteed to
+ * be <i>identical</i> to the order used by method {@link #forEachKey(DoubleProcedure)}. <p> This method can be used
+ * to iterate over the values of the receiver.
+ *
+ * @param list the list to be filled, can have any size.
+ */
+ @Override
+ public void values(IntArrayList list) {
+ list.setSize(distinct);
+ int[] elements = list.elements();
+
+ int[] val = values;
+ byte[] stat = state;
+
+ int j = 0;
+ for (int i = stat.length; i-- > 0;) {
+ if (stat[i] == FULL) {
+ elements[j++] = val[i];
+ }
+ }
+ }
+}
Propchange: lucene/mahout/trunk/math/src/main/java/org/apache/mahout/math/map/OpenDoubleIntHashMap.java
------------------------------------------------------------------------------
svn:eol-style = native