Can we get the sum or count of a column in dataiku

Hi @kusuma
,

I don't see a dataiku feature that would allow you to produce this kind of summary table. What we usually do in this case, when we are creating a final summary table, is to either export the data to an Excel file, and do the last steps with Excel; or use a python recipe to create an insight that is shown as a summary table, or directly writes an Excel file with that format.

If you want a python recipe example, let me know and I can add it here.

Cheers

Working on it!

Hi @kusuma
,

This is one option for a python code that will write the output in a Dataiku dataset:

import dataiku
import pandas as pd, numpy as np

data_for_summary = dataiku.Dataset("data_for_summary")
data_for_summary_df = data_for_summary.get_dataframe()

summarized_df = data_for_summary_df.append(
    pd.DataFrame(
        [['Grand Total', data_for_summary_df.Sessions.sum()]]
        , columns=["User Logon", "Sessions"])
)

# Write recipe outputs
summary = dataiku.Dataset("summary")
summary.write_with_schema(summarized_df)

Where the input dataset was called "data_for_summary" and the output "summary", as shown in this next screenshot:

But @NN
solution looks even better, I think!

Now, I think what DSS is currently missing is an option to create summary tables like this one using insights or the chart capabilities.

Cheers!