diff options
author | Christoph Helma <helma@in-silico.ch> | 2021-02-12 19:54:07 +0100 |
---|---|---|
committer | Christoph Helma <helma@in-silico.ch> | 2021-02-12 19:54:07 +0100 |
commit | a29eb3e38414cd252850c9c4fb356f8b2bef6fb4 (patch) | |
tree | a957d9ac455e7345c51f3ab6075698f552c497d1 /lib/dataset.rb | |
parent | 158e9a7ecbc467c3db77c354f203b1176b0fc3f2 (diff) |
model.rb refactored, mp2d models updated
Diffstat (limited to 'lib/dataset.rb')
-rw-r--r-- | lib/dataset.rb | 9 |
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 = [] |