Closed cdeil closed 7 months ago
Hey, pandas 2.0 sadly requires python 3.8. Currently, we are still supporting python 3.7, so we can't make this change yet. When we will remove support for 3.7 we can progress with this.
Do you have an ETA for dropping Python 3.7 support?
Hey, we will add support for pandas' arrow strings in the next version, due this or next week. Re-writing the internals to be more effcient, potentially using arrows - is something we want to head towards, but don't have a timeline yet.
Version 4.4.0 supports pandas strings - let me know if it worked for your usecase
Pandas now fully supports using Arrow memory and specifically Arrow strings and will make it the default in the next release: https://pandas.pydata.org/docs/whatsnew/v2.2.0.html#dedicated-string-data-type-backed-by-arrow-by-default
As mentioned in #372 we would very much appreciate memory / CPU improvements for our ADX -> pandas workloads.
Could you please add support in
azure.kusto.data.helpers.dataframe_from_result_table
or a new function to go efficiently ADX -> Arrow memory without needing to go via Python objects for strings?cc @yihezkel