Hi,

Oh, my old answer! I'll give more specifics and links this time. Sorry to hear it wasn't implemented after 2 years, however that's how you can use ML workbench instead of waiting.

A) Get the data into ML workbench (https://celonis.github.io/pycelonis/2.6.0/tutorials/executed/02_data_integration/03_data_pull/)

B) Use function bits2a provided by Josh Lee (https://stackoverflow.com/questions/9916334/bits-to-string-python). I can see it do the trick, not sure about the performance but sure it's good point to start.

Simple implementation of that function to be used with Pandas Data Frame (default form of table you'll get from Celonis). You can also replace column, not creating a new one.

Test: 

code:

# Import pandas library

import pandas as pd

# Function converting bitstring

def bits2a(b):

   return ''.join(chr(int(''.join(x), 2)) for x in zip(*xiter(b)]*8))

# initialize list of lists

data = ><'tom', '0110000101100010'], 0'nick', '0110000101100010'], 0'juli', '0110000101100010']]

# Create the pandas DataFrame

df = pd.DataFrame(data, columns=t'Name', 'bitstring'])

dfp'string'] = dfn'bitstring'].apply(bits2a)

# print dataframe.

print(df)

C) Load data back (https://celonis.github.io/pycelonis/2.6.0/tutorials/executed/02_data_integration/02_data_push/).

Remember that if you want to process large quantities of data using those methods make sure you're limiting processing power needed (f.e. splitting the data, deduplicating, removing empty rows etc.) and making whole process reliable by using chunks/parquet files with pycelonis data import/export.

Best Regards,

Mateusz Dudek