renozao / NMF

NMF: A Flexible R package for Nonnegative Matrix Factorization
139 stars 41 forks source link

Error running nmf #167

Open suminwei2772 opened 3 years ago

suminwei2772 commented 3 years ago

Hi, I'm having trouble running NMF on a large matrix of 3000 genes and 1428 samples (3000x1428 matrix). A few lines of the error I get is this:

AAML1031_rna_nmf <- nmf(log2(rna_df_most_var+1), rank=2:30, nrun=20, .pbackend=10, seed = 123456, method = 'brunet'); Timing stopped at: 0.888 0.065 0.97 Timing stopped at: 0.97 0.037 1.08 Timing stopped at: 0.903 0.178 1.083 Timing stopped at: 0.894 0.045 1.103 Timing stopped at: 0.898 0.061 1.221 Timing stopped at: 0.91 0.11 1.512 Timing stopped at: 0.997 0.099 1.504 Timing stopped at: 0.921 0.08 1.523 Timing stopped at: 0.913 0.153 1.678 Timing stopped at: 0.832 0.075 0.941 Timing stopped at: 0.877 0.042 0.996 Timing stopped at: 0.853 1.583 2.451 Timing stopped at: 0.876 0.046 1.024 Timing stopped at: 0.873 0.04 0.971 Timing stopped at: 0.845 0.093 0.943 Timing stopped at: 0.916 0.098 1.084 Timing stopped at: 0.899 0.071 1.079 Timing stopped at: 0.897 0.391 1.293 Timing stopped at: 0.89 0.018 0.975 Timing stopped at: 0.929 0.017 0.947 Timing stopped at: 36.98 9.871 9.105 Timing stopped at: 1.084 0.018 1.119 Timing stopped at: 1.088 0.049 1.268 Timing stopped at: 1.125 0.059 1.383 Timing stopped at: 1.121 0.107 1.282 I don't know what is causing the error and any help would be appreciated. Thank you, Lisa

joonan30 commented 3 years ago

This happened to me. I made it sure removing NAs and matrix for an input. It seems nrun (with parallele) is a problem.