Closed DillonHammill closed 6 years ago
It is probably due to the unsuccessful clustering from flowClust
. I've added the plot
switch so you can visualize the tmixture clustering
outcome for the diagnostic purpose. Please post on flowClust
github page for the further question regarding to the usage of flowClust
,
You generally shouldn't be clustering a linear and fluorescent channel using flowclust. The scales are so different that you'll often run into numerical issues estimating the covariance matrix.
Greg Finak
On Dec 28, 2017 11:47, "Mike Jiang" notifications@github.com wrote:
Closed #165 https://github.com/RGLab/openCyto/issues/165.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/RGLab/openCyto/issues/165#event-1403927172, or mute the thread https://github.com/notifications/unsubscribe-auth/ABUkeatLDlGkwgqI5s7DLcJZPUJkZgXZks5tE_BagaJpZM4Q48PR .
Hi Greg and Mike,
I am playing around with tmix2DGate, which seems to work well when providing two TRANSFORMED channels. When I provide a transformed channel and a linear morphology channel (e.g. FSC-A) I keep getting an error - which seems to originate in
getEllipseGate
.Thanks again for your help!