How to use dynamically filtered database in a python script

[Guilherme8482]

I'm doing an analysis that takes a subset available from my Power BI data source and generates a table using python (which performs complex calculations).

But my question is about how to bring filtered data into the Python script.

I'm using the query transformation, which has a Python script option to generate my calculation.

The problem is that when I use the filter panel (inside the panel), the Python script is never updated and the data is still wrong.

I tried to use Visual Python, but I don't want to see the Python graphs, I need to use the python data to then visualize it in the Power BI graphs.

I've looked for several solutions, but I can't get it to work.

For example, here is a code I need to run:

# The data must be filtered before applying my python script

import pandas as pd

dataset['SMA'] = dataset.VALUE.rolling(2, win_type='gaussian').sum(std=3)

# After that, I need to show using Power BI charts
 

[v-kelly-msft]

Hi  @Guilherme8482 ,

You could do the filtering in power query then load the data ,check the reference below:

https://docs.microsoft.com/en-us/power-query/filter-values

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[Guilherme8482]

Hi @v-kelly-msft ,

Thanks for the answer,

That's exactly what I'm doing, but the problem is that the power query never updated when using the python script.
Do you have any idea what I can do?

[v-kelly-msft]

Hi  @Guilherme8482 ,

Is your issue solved?

If so,could you pls mark the reply as answered to close it?

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!

[v-kelly-msft]

Hi  @Guilherme8482 ,

Even you click the "refresh" button,it cant be refreshed?

I have checked the reference,you could publish to Service and set a scheduled refresh for the dataset.Check below:

https://docs.microsoft.com/en-us/power-bi/connect-data/desktop-python-in-query-editor

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!