search

bschiffthaler / seidr

Toolkit to create crowd networks
GNU General Public License v3.0
2 stars 1 forks source link

MI bin count underflow for small matrices #26

Closed bschiffthaler closed 4 years ago

bschiffthaler commented 4 years ago

Looks like some small matrices cause the MI bin count logic to underflow. Sample log, command line and files attached:

metabolite_norm.txt metabolite_names.txt

mi -m CLR -M ~/tmp/mi.tsv -O 4 -B 24 -i /mnt/picea/projects/aspseq/nstreet/swasp/Sara/seidr_output/raw_data/metabolite_norm.tsv -g /mnt/picea/projects/aspseq/nstreet/swasp/Sara/seidr_output/raw_data/metabolite_names.tsv -o ~/tmp/clr.tsv
[ WARNING ][ 2020-04-21T17:40:24 ][ mi@microasp: 0|0 ]: Bin count 9223372036854775829 is high and might use a high amount of memory.
[ WARNING ][ 2020-04-21T17:40:24 ][ mi@microasp: 0|0 ]: Your expression matrix indicates that you have fewer genes than samples. This is rarely the case. Please make sure genes are columns, samples are rows.
[microasp:56679] *** Process received signal ***