pydata/pandas

Flexible and powerful data analysis / manipulation library for Python, providing labeled data structures similar to R data.frame objects, statistical functions, and much more

UrielMaD

@UrielMaD

Im following this docker commands

Captura de Pantalla 2020-12-01 a la(s) 14.02.33.png

Uwe L. Korn

@xhochy

@UrielMaD Where did you get that from? I cannot find it in pandas' docs.

1 reply

Still, I see that it is based of the pandas Dockerfile. I'll have a look at that.

Uwe L. Korn

@xhochy

@UrielMaD have a look at pandas-dev/pandas#38219

UrielMaD

@UrielMaD

apparently all I had to do was to create a new conda env as you said, now everything is working well

guru kiran

@gurukiran07

pd.Timedelta.isoformat has wrong output for an example in the documentation. pd.Timedelta(hours=1, seconds=10).isoformat() -> 'P0DT0H0M10S', while it should be 'P0DT1H0M10S'

Joris Van den Bossche

@jorisvandenbossche

@gurukiran07 that seems to be working correctly on pandas master:

In [6]: pd.Timedelta(hours=1, seconds=10).isoformat()
Out[6]: 'P0DT1H0M10S'

guru kiran

@gurukiran07

@jorisvandenbossche Yes, it's working fine. Example in the docs has wrong output.

3 replies

Marco Gorelli

@MarcoGorelli

I'm a bit of a licensing noob, but if I want to use scripts/generate_pip_deps_from_conda.py in another project, what does the BSD license require me to do? Do I need to just put the content of pandas/LICENSE at the top of the file?

4 replies

Irv Lustig

@Dr-Irv

The main conda repo still only has pandas 1.1.3. 1.1.5 is on conda-forge. What can we do to get the conda people to update more frequently?

Joris Van den Bossche

@jorisvandenbossche

The official way is to open issues at https://github.com/ContinuumIO/anaconda-issues/issues, but I don't know how much effect that has

The unofficial way is to have one of their employees inquire about it internally, but our direct line was @TomAugspurger, and he moved recently

Jeff Reback

@jreback

i think u can post an issue on the conda tracker as well

Irv Lustig

@Dr-Irv

I created ContinuumIO/anaconda-issues#12204 . We'll see if that gets anywhere

2 replies

Maxim Kupfer

@mbkupfer

two questions: 1. if i'm submitting a pr that requires the machine be WSL, is there a way to write a test for that? 2. what is the " whatsnew entry" ?

2 replies

Alexandre Ferreira

@alex1a_gitlab

Hi! I'm using to_csv function to export my dataframe. Anyone know how to write float64 values with %.18f precision but in exponential format (4.383400000000000074e-01)

2 replies

Ethan Miller

@ezmiller

Does anyone here know of any good resources for understanding how Pandas Indexes are implemented (i.e. the data structures they use etc.)?

3 replies

Maxim Kupfer

@mbkupfer

@jreback, I wasn't clear on your message to me re: #38546. Why do I need to open a new issue? Can't I just amend my PR?

2 replies

Jeff Reback

@jreback

yep

Erfan Nariman

@erfannariman

Not too familiar with the sampling part of numpy, but when we talk about "BitGenerator", do we mean np.random.Generator?

2 replies

Diego Torres

@DiegoAlbertoTorres

Has anyone successfully set up cygdb to debug the Pandas Cython extensions? I have tried to run build_ext as follows but I don't think it worked. cygdb was still unable to find symbols: $ python setup.py build_ext --inplace -j 4 --with-debugging-symbols --pyrex-gdb --with-cython-coverage

6 replies

Erfan Nariman

@erfannariman

Is there a reason we use a old version of numpy in our CI? I see that for some we are using 1.16.6 / 1.16.5 while the latest version is 1.19.4, see: https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=50646&view=logs&j=b516280a-8c75-541b-8ba0-ae13561718a7&t=a9217cb1-24f6-5625-717d-72085a70e1d3&l=747

Jeff Reback

@jreback

yea for backwards compatibility

we try to test oldest supported and newest

Erfan Nariman

@erfannariman

Oke thanks

Josh Friedlander

@josh-friedlander-kando

Hi, I submitted pandas-dev/pandas#38634 yesterday (with a very small change) and the CI failed. Looking at the log there's a huge amount of what looks like binary. Could all this be because of my change? Not really sure how to interpret it

Joris Van den Bossche

@jorisvandenbossche

@josh-friedlander-kando you can ignore the travis builds

(they are quite unstable lately, and we are in the process of moving away the remaining builds on travis)

So all the builds on github actions are passing, except for the linting one (but you already noted that this is due to an import order)

Josh Friedlander

@josh-friedlander-kando

oh I see, thank you!

Josh Friedlander

@josh-friedlander-kando

and one other thing - I fixed the import order, edited the commit and did git push -fto my branch. Should I cancel the PR and create a new one, or is there a way to resubmit with an edited commit? Or should I make the fix a separate commit?

4 replies

Marco Gorelli

@MarcoGorelli

Was looking at pandas-dev/pandas#32259 again

In summary: the issue is to deprecate MultiIndex.is_lexsorted as a public method and replace it with MultiIndex._is_lexsorted - users should use MultiIndex.is_monotonic_increasing.

Question is - how do we do that? I presume we need to warn about the deprecation, but how do we raise a FutureWarning (as is normally done for deprecations) without it always being triggered when the function is used internally?

Jeff Reback

@jreback

you change the internal uses to the new function

Marco Gorelli

@MarcoGorelli

oh yeah, thanks!

Markus Löning

@mloning

Hi everyone, one question about pd.DatetimeIndex and its freq attribute: why does it forget the freq information when I remove some values?

index = pd.date_range("01-01-2021", periods=10, freq="D")
index.freqstr
>>> "D"
index[[1, 2]].freqstr
>>> None

5 replies

Martin Walter

@aiwalter

Hi, I have a related question to @mloning. Why is freq=None here in the end? Is it possible to make conversion between DatetimeIndex and PeriodIndex more reliable in any way than doing like this here?

3 replies

>>> pd.date_range(start='1/1/2018 00:00:00', end='2/1/2018 00:01:00', freq="MS")
DatetimeIndex(['2018-01-01', '2018-02-01'], dtype='datetime64[ns]', freq='MS')

>>> pd.date_range(start='1/1/2018 00:00:00', end='2/1/2018 00:01:00', freq="MS").to_period().to_timestamp()
DatetimeIndex(['2018-01-01', '2018-02-01'], dtype='datetime64[ns]', freq=None)

Marco Gorelli

@MarcoGorelli

Many files have both absolute and implicit imports, e.g.


from pandas.core import ops

from .masked import BaseMaskedArray, BaseMaskedDtype

Is it desirable to make these consistent (e.g. absolute imports everywhere) or is it alright for them to be mixed?

2 replies

Erfan Nariman

@erfannariman

Is there a reason we don't have dedicated and a clear way of creating conditional columns? It's an operation you do a lot when wrangling data. For single conditions we can use loc or Series.where, but the latter does not suffice in all situations. Furthermore if you search for "pandas conditional column" you will probably end up here (https://stackoverflow.com/questions/19913659/pandas-conditional-creation-of-a-series-dataframe-column), where we can see the result of not being explicit, since there are least 10 methods mentioned. The top answer is the way to go, but still it uses two numpy methods. Also looking a bit more ahead, using np.where or np.select is not very optimal when eventually working with pandas arrays, since both methods will return a numpy array.

9 replies

jbrockmendel

@jbrockmendel

anyone else having trouble running asvs? (MacOS)

1 reply

Aidan Feldman

@afeld

:wave: new pandas contributor here. thanks to @jreback for reviewing the flurry of pull requests!

3 replies

boris

@pkarpesis:chat.avlikos.gr

[m]

Hello everyone ! Can someone help me get the desired output having these dfs :

image.png

5 replies

boris

@pkarpesis:chat.avlikos.gr

[m]

Appreciate your help !

Samuel Oranyeli

@samukweku

Hi everyone.

Could someone pls point me to the lines of code in indexing.py that takes care of slicing in .loc. so the code section that handles sth like : df.loc[:, "a":"f"]. Thanks

Marco Gorelli

@MarcoGorelli

Does anyone use the mypy daemon https://mypy.readthedocs.io/en/stable/mypy_daemon.html ? mypy pandas works fine for me, but dmypy run -- pandas throws a few errors

$ dmypy run -- pandas
pandas/_config/display.py:12: error: Need type annotation for '_initial_defencoding'  [var-annotated]
pandas/_version.py:54: error: Need type annotation for 'HANDLERS' (hint: "HANDLERS: Dict[<type>, <type>] = ...")  [var-annotated]
pandas/core/arrays/masked.py:43: error: Need type annotation for 'base'  [var-annotated]
pandas/core/computation/expressions.py:24: error: Need type annotation for '_TEST_MODE'  [var-annotated]
pandas/core/computation/expressions.py:27: error: Need type annotation for '_evaluate'  [var-annotated]
pandas/core/computation/expressions.py:28: error: Need type annotation for '_where'  [var-annotated]
pandas/core/dtypes/common.py:66: error: Need type annotation for '_is_scipy_sparse'  [var-annotated]
pandas/core/dtypes/dtypes.py:54: error: Need type annotation for 'subdtype'  [var-annotated]
pandas/core/dtypes/dtypes.py:59: error: Need type annotation for 'base'  [var-annotated]
pandas/core/generic.py:183: error: Need type annotation for '_is_copy'  [var-annotated]
pandas/core/generic.py:408: error: Need type annotation for '_ix'  [var-annotated]
pandas/core/series.py:415: error: Need type annotation for '_index'  [var-annotated]
pandas/io/excel/_base.py:748: error: Need type annotation for 'curr_sheet'  [var-annotated]
pandas/io/excel/_base.py:749: error: Need type annotation for 'path'  [var-annotated]
pandas/io/excel/_util.py:7: error: Need type annotation for '_writers' (hint: "_writers: Dict[<type>, <type>] = ...")  [var-annotated]
pandas/io/pytables.py:211: error: Need type annotation for '_table_mod'  [var-annotated]
pandas/io/sql.py:38: error: Need type annotation for '_SQLALCHEMY_INSTALLED'  [var-annotated]
pandas/plotting/_core.py:1715: error: Need type annotation for '_backends' (hint: "_backends: Dict[<type>, <type>] = ...")  [var-annotated]
pandas/plotting/_matplotlib/converter.py:43: error: Need type annotation for '_mpl_units' (hint: "_mpl_units: Dict[<type>, <type>] = ...")  [var-annotated]
pandas/tests/io/test_gbq.py:21: error: Need type annotation for 'PROJECT_ID'  [var-annotated]
pandas/tests/io/test_gbq.py:22: error: Need type annotation for 'PRIVATE_KEY_JSON_PATH'  [var-annotated]
pandas/tests/io/test_gbq.py:23: error: Need type annotation for 'PRIVATE_KEY_JSON_CONTENTS'  [var-annotated]
pandas/tseries/holiday.py:331: error: Need type annotation for 'holiday_calendars' (hint: "holiday_calendars: Dict[<type>, <type>] = ...")  [var-annotated]
pandas/tseries/holiday.py:369: error: Need type annotation for '_cache'  [var-annotated]
Found 24 errors in 16 files (checked 1179 source files)

Where communities thrive

People

Repo info

Activity