Hey-lo,

I'm building a version of weightedcalcs using conda for conda-forge. When possible, we try to include a link to the license file in the meta.yaml specification for the build; doing so requires the license be indexed in an explicit MANIFEST.in file so that it gets included in the source distribution.

This pull should add a MANIFEST.in that guarantees the license gets included, along with the change log and the Readme.

I wanted to run through your example notebook, so made this version of it which instead uses the 2016 Survey of Consumer Finances. This has the advantage of being hosted directly by the government, so anyone can run the full thing as-is (it downloads the data), e.g. in Google Colab as the above link does.

This version also removes the adult limitation, which isn't relevant at the household level (all SCF household heads are 18+).

Feel free to link it if you'd like.

Just found this package through pandas-dev/pandas#10030, really great stuff. I've put functions for working with weighted data in my microdf package, but looking to remove them in lieu of this and hopefully at some point native pandas support (MaxGhenis/microdf#55).

One feature I've found useful is the ability to pass a list of quantiles to quantile. Currently:

calc.quantile(adults, "income", [0.25, 0.75])

produces:

TypeError: '<' not supported between instances of 'list' and 'int'

This isn't that hard to do outside e.g. with list comprehension so might not be a high priority, but could save some code if it's common:

[calc.quantile(adults, "income", x) for x in [0.25, 0.75]]

Hi I run into problems when I follow the example notebook .

The section on Grouped weighted calculations fails

calc.mean(grp_marriage_sex, "income").round().astype(int)

AttributeError Traceback (most recent call last) in () ----> 1 calc.mean(grp_marriage_sex, "income").round().astype(int)

P:\Anaconda3\envs\charite\lib\site-packages\weightedcalcs\core.py in func_wrapper(self, thing, *args, **kwargs) 20 agg = thing.apply(lambda x: func(self, x, *args, **kwargs)) 21 is_series = isinstance(agg, pd.core.series.Series) ---> 22 has_multiindex = isinstance(agg.index, pd.indexes.multi.MultiIndex) 23 if is_series and has_multiindex: 24 return agg.unstack()

AttributeError: module 'pandas' has no attribute 'indexes

I am using pandas '0.20.1' and weightedcalcs '0.1.1'.

Could be a useful addition to your library. As an example, I'm interested in getting stats on race and gender in a group over time. Something like:

data_by_year = data.groupby(['year'])
race_gender_demographics = calc.distribution(data_by_year, ['race', 'gender']).round(3)

we have had this open issue in main pandas for a while: https://github.com/pandas-dev/pandas/issues/10030

here is a prospective API (which is similar to what you did): https://github.com/pandas-dev/pandas/issues/15039

If you'd have a look would be great. I think what you did here would be a nice contribution if you are interested.