Factor Analysis with Pandas and FactorAnalyzer - Real World Example

[@sumihan5036]

Thank you so much for the step-by-step procedure on performing factor analysis. I came across several python codes on it but yours is the best! For the function 'print_factor_loadings()', a value error occurs perhaps due to the version of pandas. Some suggest upgrading the version, but I solved it as follows instead of using mask(): temp_df_pos = loadings[loadings[i] >= 0.5][[i, 'variable']] Change those similar ones like it.

[@AMGaweda]

Thanks for the tip! It's been a moment since I've worked on this dataset, but since I've always been annoyed with the nested braces we need to do, it may also be the custom mask function I built up not accounting for a specific edge case. Either way, since it's one of the first things I copy into any notebook I'll have to give it a little more tinkering

[@sumihan5036]

@@AMGaweda Great! Thank you again for sharing. :)

[@chriszzz5798]

this helps a lot, thank you for sharing

[@angiez2925]

May I ask where dose the 25 come from? fa.fit=(df_fa,25) Thank you!

[@AMGaweda]

I will apologize as that's definitely a random number! The .fit() function is looking for the DataFrame (fa_df) and and the number of factors you are choosing to fit for. In this line's case, I gave it 25 (even though my data only had 11 columns). You'll notice in the "Performing Factor Analysis" section, .fit() is used more properly by saying n_factors (or 4). For your own dataset, you'd want to make this the number of features (columns) in your dataset to serve as a baseline for the 1.0 or 0.7 eigenvalue thresholds.