Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

plotting methods other than PCA in compare app #94

Open
lmsimp opened this issue May 8, 2018 · 3 comments
Open

plotting methods other than PCA in compare app #94

lmsimp opened this issue May 8, 2018 · 3 comments

Comments

@lmsimp
Copy link
Collaborator

lmsimp commented May 8, 2018

The compare app doesn't allow you to specify a plotting method. We set the plotting method internally to be PCA. I would like to change the code so we can allow other visualisations. This of course is tricky if we want to use the remap function as this only works with PCA. We could allow a call to method e.g. method = "t-SNE" only if remap = FASLE then this wouldn't break the call to remap. What do you think @lgatto

@lgatto
Copy link
Owner

lgatto commented May 9, 2018

The question is (and I think we have different opinions here) whether it makes sense to compare datasets that haven't been remapped. If we believe that some sort of remapping/normalisation is essential, then we would need to provide one before offering the visualisation. Let's discuss next week.

NB: there is no remap-like function for t-SNE (at least not in R, and I'm not sure there is one at all).

@lmsimp
Copy link
Collaborator Author

lmsimp commented May 9, 2018

Yes, re-map is straight forward with PCA and I think it should be the default.

I think it is useful however, to allow one to load two different datasets and compare locations. Does this make sense if there is no-remapping? I think the answer depends on the question and that is for the user to decide. My current use case is allowing me to look at the localisations assigned from phenoDisco and the predicted locations from curated markers. The compare app is a nice way to do this, however, I do not want to re-map and I would like to visualise with t-SNE.

@lgatto
Copy link
Owner

lgatto commented May 9, 2018

Then, I would suggest a method = c("PCA", "none") argument. When set to "none", then the two datasets are taken as they are and displayed, and it is the user's responsibility to assure that they are comparable. This needs to be highlighted in the documentation file.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants