Forum Discussion

jmarcrum's avatar
jmarcrum
Helper II
6 years ago
Solved

Slicers as input for Python script

Hi everyone -   I have a python network script visual in Power BI. I am looking to apply a slicer (instead of hard-coding the teams - tm1 and tm2 ) in the python script below. How can be this be ...
Need Help
  • TomMartens's avatar
    TomMartens
    6 years ago

    Hey jmarcrum ,

     

    this code is maybe doing what you are looking for:

    # The following code to create a dataframe and remove duplicated rows is always executed and acts as a preamble for your script: 

# dataset = pandas.DataFrame(TmName, OppName, Selection1, Selection2)
# dataset = dataset.drop_duplicates()

# Paste or type your script code here:
import networkx
import matplotlib.pyplot as plt

tm1_l = list(dataset.Selection1)
tm1 = tm1_l[1]
tm2_l = list(dataset.Selection2)
tm2 = tm2_l[1]

# load edges file from csv
left = list(dataset.TmName)
right = list(dataset.OppName)
graph = networkx.Graph()
graph.add_edges_from(list(zip(left, right)))

# Test single path
shortest_path = networkx.shortest_path(graph, tm1, tm2)

edges_2team = dataset[dataset.TmName.isin(shortest_path)]
left_2team= list(edges_2team['TmName'])
right_2team = list(edges_2team['OppName'])
graph_2team = networkx.Graph()
graph_2team.add_edges_from(list(zip(left_2team, right_2team)))
plt.title(tm1 + " " + tm2)
networkx.draw(graph_2team, with_labels=True)
plt.show()

    as it allows to create this chart:

    Be aware of this line:

    networkx.draw(graph_2team, with_labels=True)


    Hopefully, this is what you are looking for.

     

    Regards,

    Tom

TomMartens's avatar
TomMartens
Super User
6 years ago

Hey jmarcrum ,

 

you can add the column that "feeds" the slicer to the python script values.

Another option is to create a measure using DAX if the column originates from an unrelated  table. You can use SELECTEDVALUE('<tablename>'[columname] , <defaultvalue>), see here: https://dax.guide/selectedvalue/

 

Hopefully, this provides some ideas to tackle your challenge.

 

Regards,

Tom

 

Stay safe, stay healthy

    • TomMartens's avatar
      TomMartens
      Super User
      6 years ago

      Hey jmarcrum ,

       

      please provide a pbix that creates a working network graph, without trying to incorporate the unrelated table or SELECTEDVALUE solutions that I mentioned.

       

      Regards,
      Tom

      • TomMartens's avatar
        TomMartens
        Super User
        6 years ago

        Hey jmarcrum ,

         

        I couldn't resist 🙂

         

        this code

         

         

         

        # Paste or type your script code here:
import networkx
import matplotlib.pyplot as plt

tm1_l = list(dataset.Selection1)
tm1 = tm1_l[1]
tm2_l = list(dataset.Selection2)
tm2 = tm2_l[1]

# load edges file from csv
left = list(dataset.TmName)
right = list(dataset.OppName)
graph = networkx.Graph()
graph.add_edges_from(list(zip(left, right)))

# Test single path
# shortest_path = networkx.shortest_path(graph, tm1, tm2)
#
#edges_2team = dataset[dataset.TmName.isin(shortest_path)]
#left_2team= list(edges_2team['TmName'])
#right_2team = list(edges_2team['OppName'])
#graph_2team = networkx.Graph()
#graph_2team.add_edges_from(list(zip(left_2team, right_2team)))
plt.title(tm1 + " " + tm2)
networkx.draw(graph, with_labels=False)
plt.show()

         

         

         

        Creates this chart (the green rectangles are added using snagit 🙂

        As you can see the measures are working to select something.

        Please be aware that my code above is transforming the dataframe columns to a list, to select the 1st value to create the title.

        I'm using plt.title as .draw() turns off the axis, where .draw_newtworkx() is not doing this. 

         

        Hopefully, this provides some ideas about what is causing the issues

         

        Regards,

        Tom