Pandas qcut with groupby with non-unique values

Question

I'm trying to do a groupby on a pandas dataframe and on that groupby do a qcut, to classify the values on a quantile. The problem is that some groups have only one value, so qcut complains with ValueError: Bin edges must be unique . Is there a way to simply ignore these cases on the groupby and qcut?

I'm doing something like

df['quantile'] = df.groupby(['grouping'])['values'].transform(
                 lambda x: pd.qcut(x, 4))

I can do this way on a two level grouping

pd.qcut(df.groupby(['grouping', 'param1']).sum()['value'],[0.15,0.25,0.5,0.75,1.0], labels=['0.15', '0.25', '0.5', '0.75'])

But I'm not sure that the results are the quantiles inside each group for the parameter grouping or for the entire dataframe.

Answer 1

within qcut method, you could set duplicates='drop'. This should produce a small number of null values for the qcut transformation that you could decide to impute to whatever value you wish.