tarahmarie / dh-trace

MIT License
0 stars 0 forks source link

June24 tweaks (again and again and again...) #39

Closed jdmartin closed 1 month ago

jdmartin commented 1 month ago

Well, that's an entire Sunday I'll never get back...

Eltec needs to be cleaned. It has terrible, inconsistent naming and I spent my whole day playing whack-a-mole to make the SVM labels, alignment names, and names from the splits all match up. I am more than a little annoyed, and definitely done. I will not be working around Eltec again. If it doesn't need to be here, I'd recommend losing it and focusing on dissertation material.

Even so, if you want to use Eltec, you'll be using this updated alignments file (yeah, so, one of the folder names had the wrong bloody author on it!): https://drive.google.com/file/d/1AxWyivKGaszVgv5dPEdKHIU9QdiVmt94/view?usp=drive_link

(And yes, you need to run from begin.sh again. It's fun. I did it 14 times today...)

Apart from this, I have run Eltec and sm-test from scratch. I have run the make_auto_scatterplot afterwards, for both, and everything works now.

I will be leaving this to think about what it did until I calm down. At which point, there will be some cleanup to do, but nothing new or interesting apart from that.