Revisions of python-pyreadstat
- update to 1.2.7: * Fixing warnings in new pandas version 2.2.1 * Remove string encoding for read_por * Fix read in chunks when output is dict
- update to 1.2.6: * correcting bug when writing files with character columns with all missing, character length set to 1 * Correcting bug for datetime64[us * Added modification_time and creation_time to metadata * added datetime64[us in addition to ns for improvement when writing. * changes introduced for compatibility with cython 3.0.0. * added num_rows to multiprocessing to allow processing of xport, * Using os.fsencode when reading file paths to support file system file labels, reading newer por files date-like columns, and few * fixed sharing permissions when reading files on windows, - Updates for 1.0.2 * Updated default widths for DATE and DATETIME formats (from Readstat src). That make the files readable both in SPSS - Updates for 1.0.1
- update to 1.2.1: * Readstat source updated to version 1.1.9 * introduced recognition for pandas datatype datetime64[ns, UTC] and other datetime64 types when writing, so that this column type gets correctly written as datetime * introduced extra_datetime_formats and extra_date_formats arguments for read functions, cleaned the list of sas date, datetime and time formats to exclude those not directly convertible to python objects * improved performace of writer when there are datetime64 columns
Automatic submission by obs-autosubmit
- update to 1.1.4: * Included new pandas datatypes Float32Dtype and Float64Dtype when writing. * Forcing cythonization when installing from source # Updated Readstat source to version 1.1.17 * Improved error when reading dates and finding a string * repaired not reading correctly metadata with multiprocessing # correction when reading file paths in python <3.7 * Using os.fsencode when reading file paths to support file system encoding and Path-like objects * Updated Readstat source to version 1.1.6 * Accepting dict as the column_label parameter for writing functions
- Update to 1.1.0: * added meta.readstat_variable_types to track the type of binary variable in the original file. * Adding better error message for not accepting certain files types in multiprocessing, solves #105 * Fixing #106 writing failing when row 0 is removed from dataframe. * Removing numpy deprecated objects solves #107. * Adding support for pathlib.Path, soves #110. * fixing order of categories * Added formats_as_ordered_category to get an ordered category. * Added value_formats in order to be able to set the variable format when writing. - Do not build on Python 3.6, due to no NumPy.
- Update to 0.2.9: * metadata updates
Package to read and write statistical data files into pandas
Displaying all 13 revisions