Closed Bjoernsen closed 10 years ago
I think splitting out the clustering into a web worker would be an easy improvement.
I'd also like to improve the clustering algorithm which is very basic and n^2.
I'm split the clustering out into a web-worker. This seems to make the browser much more responsive even with many genes.
I loaded a data set with about 40,000 genes. The clustering and the page response were very slow.
Does it make sense to disable the heatmap if the data set has more than 3000 genes/features? Currently, the
calc_routine
is scheduled usingscheduler.schedule('heatmap.calc', calc_routine, 10*1000)
There could be a button to enable the clustering.