pandas2ri - requires pytz

[dyrvte]

Since ZoneInfo is now being used as the standard for Python I switched my pandas dataset from using pytz for timezones to ZoneInfo. It appears pandas2ri does not support this currently:

pandas2ri.py:65: UserWarning: Error while trying to convert the column ... Fall back to string conversion. The error is: 'zoneinfo.ZoneInfo' object has no attribute 'zone'

This attribute is not present in ZoneInfo

[lgautier]

@dyrvte - Can you share a minimal code example to reproduce the issue?

[dyrvte]

import rpy2.robjects as ro from rpy2.robjects import pandas2ri as p2ri from rpy2.robjects.conversion import localconverter import pandas as pd import pytz from zoneinfo import ZoneInfo from datetime import datetime #works fine record_for_prediction = pd.DataFrame([[datetime.now(),1]],columns = ["date","number"]) print(record_for_prediction) context = localconverter(ro.default_converter + p2ri.converter) with context as cv: to_predict = cv.py2rpy(record_for_prediction ) #works fine tz = pytz.UTC record_for_prediction = pd.DataFrame([[datetime.now(tz),1]],columns = ["date","number"]) print(record_for_prediction) context = localconverter(ro.default_converter + p2ri.converter) with context as cv: to_predict = cv.py2rpy(record_for_prediction ) #error occurs with this one tz = ZoneInfo("UTC") record_for_prediction = pd.DataFrame([[datetime.now(tz),1]],columns = ["date","number"]) print(record_for_prediction) context = localconverter(ro.default_converter + p2ri.converter) with context as cv: to_predict = cv.py2rpy(record_for_prediction )

…

On 30/12/2023, Laurent Gautier ***@***.***> wrote: @dyrvte - Can you share a minimal code example to reproduce the issue? -- Reply to this email directly or view it on GitHub: #1079 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>

[lgautier]

Thanks I can't reproduce on this end with rpy2 3.5.15, and pandas 2.1.3. Can you share the output of

python -m rpy2.situation

and

python -c 'import pandas; print(pandas.__version__)'

[dyrvte]

Testing with 3.5.14 and 3.5.15 on my system gives the same error. Pandas is 2.1.4, which I've just upgraded. rpy2 version: 3.5.15 Python version: 3.11.6 (main, Nov 14 2023, 09:36:21) [GCC 13.2.1 20230801] Looking for R's HOME: Environment variable R_HOME: None Calling `R RHOME`: /usr/lib64/R Environment variable R_LIBS_USER: None R's value for LD_LIBRARY_PATH: /usr/lib64/R/lib:/usr/lib/jvm/java-21-openjdk/lib/server R version: In the PATH: R version 4.3.2 (2023-10-31) -- "Eye Holes" Loading R library from rpy2: OK Additional directories to load R packages from: None C extension compilation: include: ['/usr/include/R/'] libraries: ['R', 'pcre2-8', 'lzma', 'bz2', 'z', 'tirpc', 'rt', 'dl', 'm', 'icuuc', 'icui18n'] library_dirs: ['/usr/lib64/R/lib'] extra_compile_args: ['-std=c99'] extra_link_args: ['-Wl,--export-dynamic', '-fopenmp', '-Wl,-O1,--sort-common,--as-needed,-z,relro,-z,now', '-flto=auto'] Directory for the R shared library: lib CFFI extension type Environment variable: RPY2_CFFI_MODE Value: CFFI_MODE.ANY ABI: PRESENT API: PRESENT python -c 'import pandas; print(pandas.__version__)' 2.1.4

…

On 08/01/2024, Laurent Gautier ***@***.***> wrote: Thanks I can't reproduce on this end with rpy2 3.5.15, and pandas 2.1.3. Can you share the output of ```bash python -m rpy2.situation ``` and ``` python -c 'import pandas; print(pandas.__version__)' ``` -- Reply to this email directly or view it on GitHub: #1079 (comment) You are receiving this because you were mentioned. Message ID: ***@***.***>

[lgautier]

Also working with pandas 2.1.4 on this end. I have Python 3.10.
If zoneinfo is involved it should be the same since it is available since Python 3.9 (https://docs.python.org/3/library/zoneinfo.html). I'll try with Python 3.11 later.