daru_lite 0.3.2 → 0.3.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f72d838a7ae6b15ca8d5d4d2942f733ee731815a432dec1b76e51296cd7d5fe7
-  data.tar.gz: 5653ee135b2fcf671eb28ad8af54cb11e1d8d6187bdfdef284d425b23d0da290
+  metadata.gz: c158868ddcdfad0c071d5ad8b583605e86cdd3a005714df2071d0fc122e88278
+  data.tar.gz: 56d1feb6d2b95f6a4c897589c56c32e8890fe72c9eb2e72d6b71a91ffc9e95ae
 SHA512:
-  metadata.gz: b3780b06dcb392bd866af2424ea0a19ddf263cb125b13092c3d6674ab8d6d1c576bbf6b78e6bfb794cc21aac73ac1f1af4faacb0905196c6cfb5a5190be56455
-  data.tar.gz: 4e39865f597bf9ecb63270c9f6df963ad4e8bdf6fecda6de0a305fb81b5c81b58787e28267a7e84644194a4bd63442e98fb56f642c03ef327e08b888210637ca
+  metadata.gz: 5ac43007818f83995648bdbc65c5debc0d8999c9623271f0d6ca7c70bccc99904b74ec48fd99860a38ad6d02f99493fecdffb6560cbf92fd6a6682eb1190c0a5
+  data.tar.gz: 3f980b1dd60846f9aba374663e34fbd0118396079e1366f7e0046c288087adbf0f7f606853dec915dd9b42109f30e2f25033ea277533a711f5be8ab2948e58f3

data/lib/daru_lite/data_frame/filterable.rb

@@ -107,9 +107,22 @@ module DaruLite
       end
 
       def keep_vector_if
-        @vectors.each do |vector|
-          delete_vector(vector) unless yield(@data[@vectors[vector]], vector)
+        vector_names = @vectors.to_a
+        kept_positions = @vectors.size.times.select do |position|
+          yield(@data[position], vector_names[position])
         end
+        kept_names = kept_positions.map { |position| vector_names[position] }
+
+        @data = kept_positions.map { |position| @data[position] }
+        @vectors =
+          if @vectors.is_a?(DaruLite::MultiIndex)
+            # An empty result can't form a MultiIndex, so it falls back to a plain Index.
+            kept_names.any? ? DaruLite::MultiIndex.from_tuples(kept_names) : DaruLite::Index.new(kept_names)
+          else
+            @vectors.class.new(kept_names)
+          end
+
+        self
       end
 
       # creates a new vector with the data of a given field which the block returns true

data/lib/daru_lite/version.rb

@@ -1,3 +1,3 @@
 module DaruLite
-  VERSION = '0.3.2'.freeze
+  VERSION = '0.3.3'.freeze
 end

data/spec/data_frame/filterable_example.rb

@@ -340,6 +340,53 @@ shared_examples_for 'a filterable DataFrame' do
       expect(df).to eq(DaruLite::DataFrame.new({a: [1,2,3,4,5]}, order: [:a],
         index: [:one, :two, :three, :four, :five]))
     end
+
+    context "with duplicate tuples in a MultiIndex" do
+      let(:multi_index_df) do
+        DaruLite::DataFrame.new(
+          [[1, 4], [2, 5], [3, 6]],
+          order: DaruLite::MultiIndex.from_tuples([[:count, 'A'], [:count, 'A'], [:pct, 'B']]),
+          index: %w[r1 r2]
+        )
+      end
+
+      it "keeps the matching vectors without raising on the ambiguous lookup" do
+        multi_index_df.keep_vector_if { |_vector, (_metric, label)| label != 'B' }
+
+        expect(multi_index_df.vectors).to be_a(DaruLite::MultiIndex)
+        expect(multi_index_df.vectors.to_a).to eq([[:count, 'A'], [:count, 'A']])
+      end
+
+      it "does not raise when no vector is kept" do
+        expect { multi_index_df.keep_vector_if { |_vector, _tuple| false } }.not_to raise_error
+        expect(multi_index_df.vectors.to_a).to eq([])
+      end
+    end
+
+    context "with duplicate categories in a CategoricalIndex" do
+      let(:categorical_df) do
+        DaruLite::DataFrame.new(
+          [[1, 4], [2, 5], [3, 6]],
+          order: DaruLite::CategoricalIndex.new([:a, :a, :b]),
+          index: %w[r1 r2]
+        )
+      end
+
+      it "keeps the matching vectors and preserves the CategoricalIndex" do
+        categorical_df.keep_vector_if { |_vector, category| category == :a }
+
+        expect(categorical_df.vectors).to be_a(DaruLite::CategoricalIndex)
+        expect(categorical_df.vectors.to_a).to eq([:a, :a])
+        expect(categorical_df.map_vectors(&:to_a)).to eq([[1, 4], [2, 5]])
+      end
+
+      it "preserves the CategoricalIndex when no vector is kept" do
+        categorical_df.keep_vector_if { |_vector, _category| false }
+
+        expect(categorical_df.vectors).to be_a(DaruLite::CategoricalIndex)
+        expect(categorical_df.vectors.to_a).to eq([])
+      end
+    end
   end
 
   describe "#filter_vectors" do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: daru_lite
 version: !ruby/object:Gem::Version
-  version: 0.3.2
+  version: 0.3.3
 platform: ruby
 authors:
 - Thomas Naude-Filonnière