summaryrefslogtreecommitdiff
path: root/lib/dataset.rb
diff options
context:
space:
mode:
Diffstat (limited to 'lib/dataset.rb')
-rw-r--r--lib/dataset.rb9
1 files changed, 8 insertions, 1 deletions
diff --git a/lib/dataset.rb b/lib/dataset.rb
index e30f000..f40f054 100644
--- a/lib/dataset.rb
+++ b/lib/dataset.rb
@@ -12,7 +12,7 @@ class Dataset
@independent_variable_type = File.read(File.join(@dir,"independent_variable_type")).chomp
@lines = File.readlines(file)
@header = @lines.shift.split(",")
- @header.first.match(/ID/i) ? @has_id = true : @has_id = false
+ @header.first.match(/ID|SMILES/i) ? @has_id = true : @has_id = false
@dependent_variable_name = @header.pop
@ids = []
@dependent_variables = []
@@ -43,6 +43,13 @@ class Dataset
end
@independent_variables = Matrix[ *@independent_variables ]
columns = @independent_variables.column_vectors
+ stdev = columns.collect{|c| c.to_a.standard_deviation}
+ stdev.each_index.reverse_each do |i|
+ if stdev[i] == 0
+ @independent_variable_names.delete_at(i)
+ columns.delete_at(i)
+ end
+ end
@independent_variable_means = columns.collect{|c| c.to_a.mean}
@independent_variable_standard_deviations = columns.collect{|c| c.to_a.standard_deviation}
scaled_columns = []