vecs

Procedure to vec into cluster
vecs -m /Data/tmp/twoyear -n 5 -r -l output
Use -l for feeding into cluster

————————————————–
vecs -m /Data/tmp/twoyear -n 5 -r vector output file
vecs -m (input matrix) -n (# of elements) -r (normalizes)
/Data/tmp/twoyear < >

Usage: vecs [flags]
-m Set input matrix name
-o Set output matrix name (HAL like matrix – do not do)
-d Dump to stdout (default)
-D Don’t dump matrix to stdout (don’t use; results in no output to screen or output file!)
-n Retain most variant columns (default is 9)
-M Divide cell values by cell counts
-r Give results in RSUs
-l Reverse label position (default is beginning of line)
-q converts vector elements to 0 – 1 range with .1 increments
scaled across entire matrix

for tangerine input file keep vec length <500, label at end of line
and use RSUs

*the -m simply sets the matrix to be used; the -n is the number of elements
in the
output vector; -r specifies output in RSUs

sample output:

cat 323.50 297.48 286.81 404.80 65.48
dog 71.18 262.36 344.04 486.44 121.14