Closed naikai closed 8 years ago
Added a step to check for perplexity before running t-SNE.
In default Rtsne
package, they require:
perplexity * 3 < (ncol(x) - 1)
Add checking for duplicated columns in the data Rows (genes) may have same expression values? (most likely to be those lowly expressed genes)
Some files will generate this kind of problems, such as the ALS file Wen-wei provided
Maybe add a step to check (and remove) duplicated rownames?