“I know a lot about structural variation information,” but the content of this blog post is too much, and it’s not convenient for me to look back. So I decided to split it into smaller parts and write a comprehensive one using GitBook later.

I didn’t expect to do regression again after four years… this time using R instead of SPSS. Previously, I was doing statistical analysis, but now… it’s machine learning.

Yesterday, due to some unparsable symbols in my markdown post, I thought Hexo had crashed, so I took the opportunity to switch to a new theme…and reviewed the setup process again.

After using snakemake for some time, I found that it has many practical features to facilitate daily analysis. Here is a record and整理.

I’ve never used many useful functions when writing scripts before, so I decided to learn a bit about itertools and make some notes. This time, I was dealing with large files, but most of the time, ...

R provides convenient multiprocessing capabilities, and Python has similar functionality.

Recently, I wrote a script to draw many images using ggplot. Although all the places where loops are needed have already been replaced with apply, it still can’t keep up when drawing hundreds of images at once. So, I fiddled around with parallel and managed to parallelize the plotting part to speed things up.

It’s been a while since the last update. Today, I’ll record the upgraded version of my homemade VPN!

Snakemake is indeed a very useful tool for workflow development and management. However, in certain scenarios, it can also bring some issues. Coincidentally, I discovered a rather unconventional way of using it: only use Snakemake’s dependency handling and task management, while generating scripts separately.

Writing workflows is a common task in bioinformatics analysis, and a mature and well-designed tool can greatly improve the efficiency of work.