fillna() on CAS column not working

[j-honnacker]

If I create a CAS table with missing values...

import pandas as pd
import numpy as np

df = pd.DataFrame(dict(amount=[35,40], tip=[3.5,np.nan]))

tbl = conn.upload_frame(df, casout=dict(name="test", caslib="casuser", replace=True))

...the .fillna() method does not replace the missing values:

tbl2 = tbl
tbl2['test'] = tbl2['tip].fillna(0)

tbl3 = tbl
tbl3['test'] = tbl3['tip'].fillna(0, inplace=True)

Is there any workaround? In this case, I want to add amount with tip and save the result in total_amount:

tbl['total_amount'] = tbl['amount'] + tbl['tip'].fillna(0)

[bkemper24]

as a work-around, would something like this work for what you are trying to do ?

tbl.computedVarsProgram="if tip=. then total_amount=amount; else total_amount=amount+tip;"

>>> tbl.computedVarsProgram="if tip=. then total_amount=amount; else total_amount=amount+tip;"
>>> tbl
CASTable('TEST', caslib='CASUSER', computedvarsprogram='if tip=. then total_amount=amount; else total_amount=amount+tip;')
>>> tbl.fetch()
CASResults([('Fetch', Selected Rows from Table TEST

   amount  tip  total_amount
0    35.0  3.5          38.5
1    40.0  NaN          40.0)])

[j-honnacker]

Thank you for the suggestion! Unfortunately, I was preparing a demo intended to showcase the extent to which Python syntax can be applied for profiling and cleaning a CAS table. As it often happens, what seemed simple at first (handling missing values) turned out to be (currently) unfeasible :)