summaryrefslogtreecommitdiff
path: root/training/dtrain
diff options
context:
space:
mode:
authorPatrick Simianer <p@simianer.de>2016-04-08 14:10:41 +0200
committerPatrick Simianer <p@simianer.de>2016-04-08 14:10:41 +0200
commit851f1582e839c272cbc0b6e29f5f911e281911d9 (patch)
treead6ace7d7c03fdabf4085712e5bc952cbb0f8261 /training/dtrain
parent25e5e79d469367d369f53ab694e99d9170bb11a4 (diff)
dtrain: select features explicitly
Diffstat (limited to 'training/dtrain')
-rwxr-xr-xtraining/dtrain/lplp.rb23
1 files changed, 21 insertions, 2 deletions
diff --git a/training/dtrain/lplp.rb b/training/dtrain/lplp.rb
index cf28b477..ac3fb758 100755
--- a/training/dtrain/lplp.rb
+++ b/training/dtrain/lplp.rb
@@ -1,5 +1,7 @@
#!/usr/bin/env ruby
+require 'zipf'
+
# norms
def l0(feature_column, n)
if feature_column.size >= n then return 1 else return 0 end
@@ -86,7 +88,7 @@ end
#_test()
def usage()
- puts "lplp.rb <l0,l1,l2,linfty,mean,median> <cut|select_k> <k|threshold> <#shards> < <input>"
+ puts "lplp.rb <l0,l1,l2,linfty,mean,median,/path/to/file> <cut|select_k|feature_names> <k|threshold|--> <#shards> < <input>"
puts " l0...: norms for selection"
puts "select_k: only output top k (according to the norm of their column vector) features"
puts " cut: output features with weight >= threshold"
@@ -95,8 +97,14 @@ def usage()
end
usage if ARGV.size<4
-norm_fun = method(ARGV[0].to_sym)
+norm_fun = nil
+feature_names = nil
type = ARGV[1]
+if type == 'feature_names'
+ feature_names = ARGV[0]
+else
+ norm_fun = method(ARGV[0].to_sym)
+end
x = ARGV[2].to_f
shard_count = ARGV[3].to_f
@@ -117,6 +125,17 @@ if type == 'cut'
cut(w, norm_fun, shard_count, x)
elsif type == 'select_k'
select_k(w, norm_fun, shard_count, x)
+elsif type == 'feature_names'
+ a = ReadFile.readlines_strip "#{fnames}"
+ h = {}
+ a.each { |i|
+ h[i] = true
+ }
+ w.each_pair { |k,v|
+ if h[k]
+ puts "#{k}\t#{mean(v, shard_count)}"
+ end
+ }
else
puts "oh oh"
end