You are viewing a plain text version of this content. The canonical link for it is here.
Posted to jira@arrow.apache.org by "Nic Crane (Jira)" <ji...@apache.org> on 2021/07/21 13:26:00 UTC
[jira] [Commented] (ARROW-13421) [C++] Add functionality for
reading in columns as floats from delimited files where a comma has been
used as a decimal separator
[ https://issues.apache.org/jira/browse/ARROW-13421?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17384894#comment-17384894 ]
Nic Crane commented on ARROW-13421:
-----------------------------------
Thanks [~Ruser], I have updated the ticket to request the relevant C++ functionality be added, after which we'll enable this functionality in R as well.
> [C++] Add functionality for reading in columns as floats from delimited files where a comma has been used as a decimal separator
> ---------------------------------------------------------------------------------------------------------------------------------
>
> Key: ARROW-13421
> URL: https://issues.apache.org/jira/browse/ARROW-13421
> Project: Apache Arrow
> Issue Type: Improvement
> Components: C++
> Reporter: Daniel Paierl
> Priority: Minor
>
> When reading in data where commas have been used as decimal separators (e.g. 3,141 to indicate pi), the column is read in as a character string. If I try to specify a schema in R, i.e.:
> {{tbl <- tibble::tibble(x = rnorm(5))}}
> {{# write to disk with comma separator}}
> {{readr::write_csv2(tbl, "tst.csv")}}
> {{# read back in}}
> {{read_delim_arrow("tst.csv", delim = ";", schema = schema(x = float32()))}}
> I get the following error:
> {{Error: Invalid: In CSV column #0: CSV conversion error to float: invalid value 'x'}}
> {{/home/nic2/arrow/cpp/src/arrow/csv/converter.cc:437 decoder_.Decode(data, size, quoted, &value)}}
> {{/home/nic2/arrow/cpp/src/arrow/csv/parser.h:84 status}}
> {{/home/nic2/arrow/cpp/src/arrow/csv/converter.cc:441 parser.VisitColumn(col_index, visit) }}
> Please can we have the functionality to be able to read in data from this format as it's fairly common across a number of countries?
--
This message was sent by Atlassian Jira
(v8.3.4#803005)