Processed 116v/17r Multispectral

[npcompl33t]

I do appreciate this thread, and I genuinely hope what you are posting is correct, but this is the reason that heavily AI-assisted analysis and responses are supposed to be against the rules here. I'm not really talking to anybody, and you are just giving me arxiv links to general broad concepts and stating that's how you did it. You're acting as if a link to mathworks or an arxiv paper would somehow set me up to do what you have done. 

What are you looking for here exactly? I shared the exact methods -- including code to reproduce this exactly. PCA is a standard analysis, that is discussed all the time on this forum. However it is something you would learn in an advanced undergraduate or graduate level class. It involves a lot of linear algebra, its beyond the scope of this post for to explain, you are just going to have to take me word for it or look it up yourself. 

The other papers i shared had nothing to do with how the PCA3 image was generated. They were in response to two points being raised:

• The marks represent some kind of correction or different ink
  
• the arxiv paper links were methods for identifying faded writing in manuscripts, specifically to look if some charcters were p's or had descenders. It had nothing to do with the section you were commetning on, it was something I noticed and raised seperately. 
  

I didn't independently test the published methods, but they directly applied to the questions that were being raised. I'm not arguing or debating their validity. they have noting to do with how i created the PCA3 image, or your question. 

I was writing those posts while you responded, they were not in response to you. This is the nature of having a forum that does not have thread replies. I replied directly to you with a succinct explanation, and attached python code so that you can generate it yourself.