Nima Sarajpoor

That is correct and it is aligned with the concept of replacing a constant subsequence with all 0. I will provide a link to its proof.

@seanlaw I tried to see if I can open the notebook in incognito mode (like when you want to open it on your end), and I couldn't. So, I submitted...

@seanlaw And, one more thing: it would be nice if someone can read the paper and see if the authors investigate this matter further or not. I can read it...

@seanlaw **Update (might be of your interest)** I took a look at [the paper](https://core.ac.uk/download/pdf/287941767.pdf). It seems it is about handling noise in time series data and how it affects the...

@seanlaw I see. I am trying to figure out "HOW / WHY" this concept should be used when we are dealing with constant / almost constant subsequences. ~~So, for instance,...

**A few notes** * I derived `eq(7)` of paper, and it is correct. * The formula in Algorithm 1 has a typo. The correct one is provided in the section...

Thanks for the clarification. I was actually trying to see if we can borrow an idea from the paper but I think I should better not go to that route....

@rasbt I just noticed that this PR was initially opened in 2019! I did a review. It would be great if you could take a look and see if they...

> Yes, `fixed_features` are already in the `SequentialFeatureSelector` but not in the `ExhaustiveFeatureSelector`, and it would be awesome to have it. Good to know! I didn't know this feature is...

@rasbt You may want to close this.