Python Visual built using data from two tables, no relationship. Any workarounds?

[PBI-Beginner]

Hi, 

I understand PowerBI Python visuals have a limitations when it comes to plotting from two different datasets with no relationships. However I was wondering if there could be a workaround in my case.

I want to plot a PCA scatter plot with product data and a static confidence ellipse with static product training data (this data must not be affected by slicers). Then have a date slicer that I can filter the product data by date to see how many of my product points fall within the static ellipse and if I have any outliers.

Product Data (sample):

Date	PC1	PC2
07-05-20	0.23	0.25
08-05-20	0.28	0.29

Training Data (sample)

PC1_Train	PC2_Train
0.24	0.26
0.28	0.22

Is there any way to build that static ellipse from the training data table and without having to define any relationship? I thought of including the training columns in my Product Data column but then the date slicer affects the Training Data, effectively changing the ellipse everytime I slice making the comparisons difficult 

Is there any workaround that I can explore or do I just have to get rid of the date slicer (which will be a major feature loss).

Extract of my python visual code:

plt.scatter(a, b) #Scatter plot with product data
plt.plot(c * np.cos(t), d * np.sin(t) , color = 'black') #Plot Ellipse with training data

#Where a and b are columns with Product PC1 and PC2 data which is expected to be sliced
#C and D are linked to PC1 and PC2 Training data currently from a different table

WIll appreciate any help.

[PBI-Beginner]

I want to plot a PCA scatter plot with product data and a static confidence ellipse with static product training data (this data must not be affected by slicers). Then have a date slicer that I can filter the product data by date to see how many of my product points fall within the static ellipse and if I have any outliers.

Product Data (sample):

Date	PC1	PC2
07-05-20	0.23	0.25
08-05-20	0.28	0.29

Training Data (sample)

PC1_Train	PC2_Train
0.24	0.26
0.28	0.22

Is there any way to build that static ellipse from the training data table and without having to define any relationship? I thought of including the training columns in my Product Data column but then the date slicer affects the Training Data, effectively changing the ellipse everytime I slice making the comparisons difficult 

Is there any workaround that I can explore or do I just have to get rid of the date slicer (which will be a major feature loss).

Extract of my python visual code:

plt.scatter(a, b) #Scatter plot with product data
plt.plot(c * np.cos(t), d * np.sin(t) , color = 'black') #Plot Ellipse with training data

#Where a and b are columns with Product PC1 and PC2 data which is expected to be sliced
#C and D are linked to PC1 and PC2 Training data currently from a different table

WIll appreciate any help.

[v-kelly-msft]

Hi @PBI-Beginner ,

Pls refer to :

https://radacad.com/userelationship-or-role-playing-dimension-dealing-with-inactive-relationships-in...

Best Regards,
Kelly

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

[PBI-Beginner]

Hi @v-kelly-msft ,

Thanks for your reply.

I tried creating an inactive relationship between the two tables as a workaround but the python visual refused to work, demanding for an active relationship.

In my case, I need a static confidence ellipse to be created from two columns on table 2, and that data (on table 2) must not be affected by any slicers or filters applied to data on table 1 (hence the inactive relationship).

Regards,