willwerscheid flashier issues

willwerscheid / flashier

A faster and angrier package for EBMF.

https://willwerscheid.github.io/flashier/

Other

10 stars 12 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

init.tol is too loose for large problems

#36 willwerscheid closed 5 years ago
1
use warmstarts for pn greedy

#35 willwerscheid closed 2 years ago
2
Propagate data names to loadings.psd and loadings.lfsr

#34 willwerscheid closed 2 years ago
1
Extrapolate

#33 willwerscheid closed 5 years ago
0
Investigate problems with NN fits

#32 willwerscheid closed 2 years ago
2
"dropout" backfit order shouldn't be default

#31 willwerscheid closed 5 years ago
0
add elbo as output option

#30 willwerscheid closed 5 years ago
0
Change name of greedy.tol

#29 willwerscheid closed 5 years ago
1
Suppress warnings about ignoring mode for nonzero.mode prior families

#28 willwerscheid closed 5 years ago
0
Suppress ash warnings about non-convergence

#27 willwerscheid closed 5 years ago
0
Replace backfit parameter with "method" parameter

#26 willwerscheid closed 5 years ago
1
"noisy" variance estimation is very slow

#25 willwerscheid opened 5 years ago
0
interface changes

#24 willwerscheid closed 5 years ago
3
allow low-rank Y

#23 willwerscheid closed 5 years ago
0
add "normal" prior.type

#22 willwerscheid closed 5 years ago
3
add "get fitted values" to interface

#21 willwerscheid closed 5 years ago
0
In ebnm.ash you are passing argument "output" to ash

#20 pcarbo closed 5 years ago
0
add restart.after.nullcheck parameter

#19 willwerscheid closed 5 years ago
0
add nullcheck threshold

#18 willwerscheid closed 5 years ago
0
bug in lowrank.square

#17 willwerscheid closed 5 years ago
1
Need to skip nullcheck when there aren't any factors

#16 willwerscheid closed 5 years ago
0
change greedy.Kmax logic

#15 willwerscheid closed 5 years ago
0
add lfsr to returned object

#14 willwerscheid closed 5 years ago
1
fixed factors interface is really clunky

#13 willwerscheid closed 5 years ago
0
allow data to be Matrix object even when S is not

#12 willwerscheid closed 5 years ago
1
fix n.factors logic

#11 willwerscheid closed 5 years ago
1
nonnegative initialization issues

#10 willwerscheid closed 5 years ago
0
allow data = NULL when flash.init is not null

#9 willwerscheid closed 5 years ago
0
fix bug in sampler when nullcheck removes factors

#8 willwerscheid closed 5 years ago
0
When possible, use Rcpp to do initial calculation of R2

#7 willwerscheid opened 5 years ago
0
replace is with inherits

#6 willwerscheid closed 5 years ago
0
Fix ref to flashr in workhorse default; change way output is obtained from ash

#5 willwerscheid closed 5 years ago
1
Check no zero rows/cols in data matrix

#4 willwerscheid closed 5 years ago
0
Deal with zero cols when normalizing loadings

#3 willwerscheid closed 5 years ago
0
Handle convergence issues more informatively

#2 willwerscheid opened 5 years ago
1
idea for "montaigne"

#1 willwerscheid closed 5 years ago
1