search

LTLA / bluster

Clone of the Bioconductor repository for the bluster package.
https://bioconductor.org/packages/devel/bioc/html/bluster.html
2 stars 3 forks source link

RMSD not taking root mean square but sum of squares #10

Closed davnovak closed 3 years ago

davnovak commented 3 years ago

Thanks for the package!

Hope I'm not misinterpreting, but in the function clusterRMSD, it seems you're not in fact computing the root mean square, you're just summing square deviations.

LTLA commented 3 years ago

Oops. Forgot a sqrt() somewhere in the last-minute refactoring.

Note that the square root of the sum of the variances is the RMSD, as the "mean" refers to that across cells in the same cluster (i.e., the expectation of the squared deviations), rather than across dimensions. At least, that's how I interpreted it. It doesn't really matter either way; if you like the other interpretation, you can get from one to the other by scaling by sqrt(<no. of dims>).

LTLA commented 3 years ago

Closed by 1358d6e3394ba0313dc002a97f9736d5b62d98b0, bluster v1.2.1 should hit the shelves soon.