Closed Marius1311 closed 1 year ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
Hi @michalk8, I still have to add this new pancreas dataset version to figshare and change this to use myst, but otherwise this is finally ready for your review.
Test are failing due to scVelo, I partially addressed this in https://github.com/theislab/scvelo/pull/1059, but https://github.com/theislab/scvelo/issues/1044 is still not covered.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:26Z ----------------------------------------------------------------
I wouldn't give this a priority over the paragraph below.
Marius1311 commented on 2023-06-13T14:18:57Z ----------------------------------------------------------------
makes sense. I moved it.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:27Z ----------------------------------------------------------------
Headings are too high, should be lower than Computing initial and terminal states
This holds also in the CytoTRACE tutorial and the Pseudotime tutorial.
Marius1311 commented on 2023-06-13T14:19:11Z ----------------------------------------------------------------
yes, I'll change this everywhere.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:28Z ----------------------------------------------------------------
The sentence "This tutorial does not depend" is too long, would split at "you may"
Would also rephrase (from negative to positive) and mention that this tutorial works with every other kernel + highlight this in bold.
Marius1311 commented on 2023-06-13T14:24:54Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:29Z ----------------------------------------------------------------
you will need... -> you will need:
Please link to the other tutorials when mentioning the other kernels
Marius1311 commented on 2023-06-13T14:29:15Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:30Z ----------------------------------------------------------------
Would move to the note where you discuss the scverse forum.
Marius1311 commented on 2023-06-13T14:29:38Z ----------------------------------------------------------------
I moved it to the first paragraph.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:31Z ----------------------------------------------------------------
Line #4. from scipy.stats import ttest_ind
Would slighly prefer import scipy.stats as st
ok, adapted.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:31Z ----------------------------------------------------------------
Use ~
when referring to anndata.AnnData
.
"When you use a kernel to compute a transition matrix, this matrix, as well as a few informations about the computation, are written "
^ is not completely true, you have to call write_to_adata
, maybe mention this.
Marius1311 commented on 2023-06-13T15:17:21Z ----------------------------------------------------------------
done, thanks!
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:32Z ----------------------------------------------------------------
The {mod}~cellrank.estimators
allowo you to analyze cell-state dynamics. We initialize the GPCCA estimator...
However, this and other estimators with with every CellRank {mod}kernel <cellrank.kernels>
.
Marius1311 commented on 2023-06-13T15:19:53Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:33Z ----------------------------------------------------------------
as we explain -> as explained
Marius1311 commented on 2023-06-13T15:20:53Z ----------------------------------------------------------------
well...
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:34Z ----------------------------------------------------------------
"We'll demonstrate both modes in this tutorial."
There's a way to add links to section headers in myst, see, e.g., here: https://myst-parser.readthedocs.io/en/latest/syntax/cross-referencing.html#implicit-targets
{meth}schur decomposition
-> {meth}Schur decomposition
Marius1311 commented on 2023-06-13T15:26:19Z ----------------------------------------------------------------
ok, done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:35Z ----------------------------------------------------------------
Add links to fit()
and predict()
using the {meth} role.
Marius1311 commented on 2023-06-13T15:26:52Z ----------------------------------------------------------------
I removed these from the section heading.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:36Z ----------------------------------------------------------------
Two spaces in "algorithm as"
as implemented in the pyGPCCA package (cite the zenodo)
Marius1311 commented on 2023-06-13T15:27:43Z ----------------------------------------------------------------
included already (reuter:22)
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:38Z ----------------------------------------------------------------
CellRank corases-grains -> CellRank coarse-grains.
If I remember correctly, cellrank.estimators.GPCCA.macrostates
is the hard assignment,cellrank.estimators.GPCCA.macrostates_memberships
(please check) is the soft assignment.
Marius1311 commented on 2023-06-13T15:31:04Z ----------------------------------------------------------------
you're right, that seems to be the case. adapted.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:39Z ----------------------------------------------------------------
Can we please remove the horizontal and vertical lines + fix the figure aspect ratio?
Marius1311 commented on 2023-06-13T15:46:12Z ----------------------------------------------------------------
I don't know how
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:40Z ----------------------------------------------------------------
adata.obs['clusters']
-> {attr}adata.obs['clusters'] <anndata.AnnData.obs>
.
{mod}pyGPCCA
doesn't link to anything
for more infos -> for more information.
Marius1311 commented on 2023-06-13T15:47:47Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:40Z ----------------------------------------------------------------
When mentioning PETSc/SLEPc, would consider adding the conda install -c conda-forge petsc4py slepc4py
Please don't hardcode the link to installation instructions, use the proper {doc}.
Marius1311 commented on 2023-06-13T15:50:48Z ----------------------------------------------------------------
ok, done
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:42Z ----------------------------------------------------------------
"we use the ... method to identify the terminal states."
Marius1311 commented on 2023-06-13T15:51:24Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:43Z ----------------------------------------------------------------
continious -> continuous
Marius1311 commented on 2023-06-13T15:52:01Z ----------------------------------------------------------------
ok, done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:44Z ----------------------------------------------------------------
Maybe mention higher intensity/saturation means higher certainty.
Marius1311 commented on 2023-06-13T15:54:18Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:45Z ----------------------------------------------------------------
states using the {meth}...
method.
Marius1311 commented on 2023-06-13T15:55:41Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:46Z ----------------------------------------------------------------
"and can be inspected conveniently."
I'd have expected you should g.initial_states
and g.terminal_states
Marius1311 commented on 2023-06-13T16:01:44Z ----------------------------------------------------------------
Ok, I added that.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:47Z ----------------------------------------------------------------
inintial -> initial
Marius1311 commented on 2023-06-13T16:02:45Z ----------------------------------------------------------------
thanks.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:48Z ----------------------------------------------------------------
initial- -> initial
Marius1311 commented on 2023-06-13T16:03:01Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:49Z ----------------------------------------------------------------
Would not use italics for terminal
.
knowlege -> knowledge
which we show below -> as shown below.
Marius1311 commented on 2023-06-13T16:03:44Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:50Z ----------------------------------------------------------------
probabilits -> probability
Marius1311 commented on 2023-06-13T16:04:09Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:51Z ----------------------------------------------------------------
we compute the macrostates.
dashed vertical line
Marius1311 commented on 2023-06-13T16:04:49Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:52Z ----------------------------------------------------------------
Remove the Nice!
...; however, that does -> . However, that does not
Missing .
in the last sentence.
; we show below -> . We show below
" for each of the cells most confidently assigned to a certain macrostates" reads not so easily
Marius1311 commented on 2023-06-13T16:06:30Z ----------------------------------------------------------------
thanks, done
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:53Z ----------------------------------------------------------------
any {attr}...
-> any observation-level covariate
Link to the transport map kernel tutorial rather than the class (or both) (also missing the {class} role for the kernel).
Marius1311 commented on 2023-06-13T16:09:29Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:54Z ----------------------------------------------------------------
becasue -> because
Let's try the automatic terminal
Marius1311 commented on 2023-06-13T16:10:54Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:54Z ----------------------------------------------------------------
ductal-cell markers -> Ductal cell markers.
Marius1311 commented on 2023-06-13T16:11:44Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:55Z ----------------------------------------------------------------
as initial -> as an initial state.
Marius1311 commented on 2023-06-13T16:12:02Z ----------------------------------------------------------------
ok.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:10:56Z ----------------------------------------------------------------
we recommend... -> we recommend to: + go thrgouh/take a look/take a look/
Marius1311 commented on 2023-06-13T16:12:46Z ----------------------------------------------------------------
thanks!
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:15:06Z ----------------------------------------------------------------
{attr) -> {attr}
Marius1311 commented on 2023-06-13T16:10:10Z ----------------------------------------------------------------
done.
View / edit / reply to this conversation on ReviewNB
michalk8 commented on 2023-06-13T01:15:07Z ----------------------------------------------------------------
compute the {attr}fate probabilities <...>
.
Marius1311 commented on 2023-06-13T16:11:19Z ----------------------------------------------------------------
done.
This adds a new tutorial for initial and terminal state identification.