Paul Brodersen

Yeah, that exact example motivated the qualifier: > More often than not, [...] However, the fact that we had the exact same example in mind is maybe telling that there...

I have no citations just anecdotes (which I would attribute to classic survivorship bias). For example, I was once approached by a startup to develop them a neural net that...

I think this need a corollary: Normalize, normalize, normalize (...but don't normalize your data prematurely). I think every data scientist goes through a phase were they just throw some sort...

I also seem to get the same problem with `IndependentComponentsDistribution` (using `NormalDistribution` for the different dimensions).

I am just glad the bug is reproducible. It was driving me absolutely mad when it happened.

> Now that I have this retraining procedure, I added back in the plotting code and I'm able to converge on some trained model, so I don't know if this...

I know this issue is old, but you could use my [matplotlib-venn-wordcloud](https://github.com/paulbrodersen/matplotlib_venn_wordcloud) to spread the text around in a form filling manner (that's why I wrote that little module to...

Sure, can do. Apart from the code example above, what else would you like to see in the notebook?

1. What is an orthogonal layout? 2. To answer your question: no, the only layout algorithm that I have implemented for the time being is Fruchtermann-Reingold. I do have plans...

Is [this](https://arxiv.org/abs/1807.09368) what you are referring to?