v2022.03.15

Highlights

• Lots of new functionality from akutil, including DataFrames, Series, Index, and SegArray
• New grouping API
• Lots of parquet improvements, including error handling and string read performance

What's Changed

• Closes #1110 legacy_placeholder removal by @Ethan-DeBandi99 in #1123
• more variations on the arkouda logo by @mhmerrill in #1143
• Fix quoting when getting vars from the Makefile by @ronawho in #1148
• Fix Parquet unit test version number test when Arrow version < 5 by @bmcdonald3 in #1150
• Fix crash when passing in bad column names to Parquet by @bmcdonald3 in #1157
• Remove trailing space for ak.get_datasets by @bmcdonald3 in #1159
• Closes #1121 uint64 hash return by @Ethan-DeBandi99 in #1147
• Closes #1145: Add uint64 support for ak.arange by @pierce314159 in #1152
• Closes #1012 .astype() method by @Ethan-DeBandi99 in #1160
• Change ak.read_parquet to read all supported datasets by default by @bmcdonald3 in #1146
• Closes #1118 move segarray/dataframe from akutil by @Ethan-DeBandi99 in #1139
• Add support for reading and writing bool Parquet columns by @bmcdonald3 in #1156
• Add support for reading string columns from Parquet files by @bmcdonald3 in #1163
• Add support for reading and writing float64 Arrow columns by @bmcdonald3 in #1170
• Closes #1126 dtypes and util to arkouda by @Ethan-DeBandi99 in #1171
• Closes #1127 alignment to arkouda by @Ethan-DeBandi99 in #1172
• Get Parquet string reading offsets in C++ function by @bmcdonald3 in #1175
• Add UsedModules.cfg to .gitignore by @bmcdonald3 in #1179
• Add line to pass Arrow float to Chapel side by @bmcdonald3 in #1186
• Closes #1165 drop on axis by @Ethan-DeBandi99 in #1177
• Update Arkouda to work with CTypes by @bradcray in #1185
• Catch all Parquet errors and report them to the client by @bmcdonald3 in #1188
• Closes #1181: use apache/parquet-testing files by @reuster986 in #1183
• Update Arrow version to 7 and add snappy to install by @bmcdonald3 in #1196
• Add support for reading binary data columns in Parquet by @bmcdonald3 in #1195
• Closes #1176 doc version num by @Ethan-DeBandi99 in #1198
• Closes #1022 collapse unique and value_counts server messages by @Ethan-DeBandi99 in #1192
• Issue #1189: Fix type of H5Literate argument to H5_iter_order_t by @glitch in #1201
• Closes #1128 series/index into arkouda by @Ethan-DeBandi99 in #1182
• Closes #1155: ak.cast doesn't work with numpy dtypes by @pierce314159 in #1202
• Closes #1178 ak.DataFrame.from_pandas() by @Ethan-DeBandi99 in #1184
• Only read necessary elements in Chapel array for Parquet by @bmcdonald3 in #1204
• Closes #1207: Implement ak.full and ak.full_like by @pierce314159 in #1208
• Update optional Parquet test to account for string and binary changes by @bmcdonald3 in #1205
• Closes #1203 - SegArray - Allow Empty Segments by @Ethan-DeBandi99 in #1206
• Closes #1164 ak.DataFrame Testing Adjustments by @Ethan-DeBandi99 in #1191
• Consolidate and standardize grouping API by @reuster986 in #1212

Full Changelog: v2022.02.23...v2022.03.15