do you tend to put the year that a script was created or when it was last updated?
I typically sign my name and date upon script creation. However, since github tracks these types of things you can easily see contributions (when and who), so it doesn't really matter.
I wonder if you might want to add additional details in your function for "apply_signature" or add a comment where you all this function because I'm not exactly sure if i'm understanding the output correctly.
Yeah, that sounds like a good idea. I added documentation to the function definition.
Based on the function call it looks like you're returning the vae_111 feature of your nbl_df but I'm not sure how you got this.
Yep! In the function i specify top_feature (which stores the string vae_111). Basically I multiply nbl_df by the compressed feature vae_111
When you are retrieving the weight matrix, I was expecting to see vae features, instead there are pca features. Is this correct?
Yeah, the data structure decision I made early on was to save all weight matrices (within a specific k) in the same matrix. So that weight matrix has PCA, ICA, NMF, DAE, and VAE features.
I typically sign my name and date upon script creation. However, since github tracks these types of things you can easily see contributions (when and who), so it doesn't really matter.
Yeah, that sounds like a good idea. I added documentation to the function definition.
Yep! In the function i specify
top_feature
(which stores the stringvae_111
). Basically I multiplynbl_df
by the compressed featurevae_111
Yeah, the data structure decision I made early on was to save all weight matrices (within a specific
k
) in the same matrix. So that weight matrix has PCA, ICA, NMF, DAE, and VAE features.