search

Talavera-Lopez-Lab / COPD_influenza

MIT License
0 stars 0 forks source link

Comp analysis #2

Closed egerc closed 8 months ago

review-notebook-app[bot] commented 11 months ago

Check out this pull request on  ReviewNB

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:17Z ----------------------------------------------------------------

You can use hyperlinks to reduce the space that weblinks take. For example, best practicesl tutorial


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:18Z ----------------------------------------------------------------

I like a lot that you are documenting troubleshooting. This is really nice. I would just add standard Markdown format to make it more organised.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:19Z ----------------------------------------------------------------

You don't need the print(adata) part here. You can just write adata and the same info would be displayed.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:20Z ----------------------------------------------------------------

This usually takes too much space and make the notebooks unecessarily large. You could just check the first lines with adata.obs.head() which will work because this is a data frame.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:21Z ----------------------------------------------------------------

You need a header here. It is important to separate the sections of your analysis, so people can follow your code.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:21Z ----------------------------------------------------------------

1) I don't really like subcomments (The ones starting with #) inside each cell. But that's my personal taste.

2) Here you have three steps in a single code cell. I would prefer if you split it in three separate ones: one for the log counts preparation, one for the HVGs, and one for the PCA.

3) PCA, Neighbours, and UMAP are not deterministic, so we need to add a random_seed= parameter for all of them. My favourite is random_seed=1712 .

Questions:

  • Why did you select 3K genes?
  • Why only 10 neighbours and 30 PCs?


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:22Z ----------------------------------------------------------------

Nice!


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:23Z ----------------------------------------------------------------

Here, you are using 50 neighbours, but before, you used 10. Why?


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:24Z ----------------------------------------------------------------

Add a brief explanation of what this means.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:16:25Z ----------------------------------------------------------------

This section needs also a header, and a brief explanation of why you are selecting this model parameters.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:42:02Z ----------------------------------------------------------------

The function needs to go at the top, after the section where you set up the environment.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:42:03Z ----------------------------------------------------------------

The commented code is OK while you are optimising and testing. But by the time you commit a PR or final version, it has to be removed.


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:42:04Z ----------------------------------------------------------------

Beautiful!


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:42:05Z ----------------------------------------------------------------

Same as above


review-notebook-app[bot] commented 9 months ago

View / edit / reply to this conversation on ReviewNB

cartal commented on 2024-01-14T15:42:06Z ----------------------------------------------------------------

If this didn't work and you don't need it, remove it.


cartal commented 9 months ago

We still need to do some corrections before merging.

  • © Githubissues.
  • Githubissues is a development platform for aggregating issues.