davidrichards-data_frame 0.0.18 → 0.0.19
Sign up to get free protection for your applications and to get access to all the features.
- data/README.rdoc +16 -0
- data/VERSION.yml +1 -1
- data/bin/plain_frame +22 -0
- data/lib/data_frame.rb +2 -1
- data/lib/data_frame/arff.rb +43 -36
- data/lib/data_frame/core/column_management.rb +102 -0
- data/lib/data_frame/core/filter.rb +48 -0
- data/lib/data_frame/core/import.rb +112 -0
- data/lib/data_frame/core/pre_process.rb +61 -0
- data/lib/data_frame/core/saving.rb +29 -0
- data/lib/data_frame/core/training.rb +36 -0
- data/lib/data_frame/data_frame.rb +37 -241
- data/lib/data_frame/id3.rb +28 -0
- data/lib/data_frame/kmeans.rb +10 -0
- data/lib/data_frame/labels_from_uci.rb +48 -0
- data/lib/data_frame/mlp.rb +18 -0
- data/lib/data_frame/sbn.rb +18 -0
- data/lib/data_frame/transposable_array.rb +1 -1
- data/lib/ext/array.rb +11 -0
- data/spec/data_frame/arff_spec.rb +1 -0
- data/spec/data_frame/core/column_management_spec.rb +97 -0
- data/spec/data_frame/core/filter_spec.rb +88 -0
- data/spec/data_frame/core/import_spec.rb +41 -0
- data/spec/data_frame/core/pre_process_spec.rb +71 -0
- data/spec/data_frame/core/saving_spec.rb +61 -0
- data/spec/data_frame/core/training_spec.rb +51 -0
- data/spec/data_frame/data_frame_spec.rb +10 -226
- data/spec/data_frame/id3_spec.rb +22 -0
- data/spec/ext/array_spec.rb +13 -0
- data/spec/fixtures/discrete_testing.csv +4 -0
- data/spec/fixtures/discrete_training.csv +21 -0
- metadata +33 -6
@@ -21,12 +21,6 @@ describe DataFrame do
|
|
21
21
|
@df.items.should be_empty
|
22
22
|
end
|
23
23
|
|
24
|
-
it "should be able to add an item" do
|
25
|
-
item = [1,2,3,4]
|
26
|
-
@df.add_item(item)
|
27
|
-
@df.items.should eql([item])
|
28
|
-
end
|
29
|
-
|
30
24
|
it "should use just_enumerable_stats" do
|
31
25
|
[1,2,3].std.should eql(1)
|
32
26
|
lambda{[1,2,3].cor([2,3,5])}.should_not raise_error
|
@@ -74,12 +68,15 @@ describe DataFrame do
|
|
74
68
|
end
|
75
69
|
end
|
76
70
|
|
77
|
-
it "should be able to
|
78
|
-
|
79
|
-
|
80
|
-
|
81
|
-
|
82
|
-
@
|
71
|
+
it "should be able to initialize from an array" do
|
72
|
+
contents = %{7,5,mar,fri,86.2,26.2,94.3,5.1,8.2,51,6.7,0,0
|
73
|
+
7,4,oct,tue,90.6,35.4,669.1,6.7,18,33,0.9,0,0
|
74
|
+
}
|
75
|
+
|
76
|
+
@labels = [:x, :y, :month, :day, :ffmc, :dmc, :dc, :isi, :temp, :rh, :wind, :rain, :area]
|
77
|
+
@df = DataFrame.new(@labels)
|
78
|
+
@df.import(contents)
|
79
|
+
@df.labels.should eql(@labels)
|
83
80
|
end
|
84
81
|
|
85
82
|
context "csv" do
|
@@ -89,7 +86,7 @@ describe DataFrame do
|
|
89
86
|
7,4,oct,tue,90.6,35.4,669.1,6.7,18,33,0.9,0,0
|
90
87
|
}
|
91
88
|
labels = [:x, :y, :month, :day, :ffmc, :dmc, :dc, :isi, :temp, :rh, :wind, :rain, :area]
|
92
|
-
|
89
|
+
|
93
90
|
@df = DataFrame.from_csv(contents)
|
94
91
|
@df.labels.should eql(labels)
|
95
92
|
@df.x.should eql([7,7])
|
@@ -109,22 +106,6 @@ describe DataFrame do
|
|
109
106
|
end
|
110
107
|
end
|
111
108
|
|
112
|
-
it "should be able to remove a column" do
|
113
|
-
@df = DataFrame.new :twos, :threes, :fours
|
114
|
-
@df.import([[2,3,4], [2,3,4], [2,3,4], [2,3,4]])
|
115
|
-
@df.drop!(:twos)
|
116
|
-
@df.items.all? {|i| i.should eql([3,4])}
|
117
|
-
@df.labels.should eql([:threes, :fours])
|
118
|
-
end
|
119
|
-
|
120
|
-
it "should be able to remove more than one column at a time" do
|
121
|
-
@df = DataFrame.new :twos, :threes, :fours
|
122
|
-
@df.import([[2,3,4], [2,3,4], [2,3,4], [2,3,4]])
|
123
|
-
@df.drop!(:twos, :fours)
|
124
|
-
@df.items.all? {|i| i.should eql([3])}
|
125
|
-
@df.labels.should eql([:threes])
|
126
|
-
end
|
127
|
-
|
128
109
|
it "should offer a hash-like structure of columns" do
|
129
110
|
@df.add [1,2,3,4]
|
130
111
|
@df.add [5, 6, 7, 8]
|
@@ -156,202 +137,5 @@ describe DataFrame do
|
|
156
137
|
@df.variables.should eql(@labels)
|
157
138
|
end
|
158
139
|
|
159
|
-
context "replace!" do
|
160
|
-
before do
|
161
|
-
@df.add [1,2,3,4]
|
162
|
-
@df.add [5, 6, 7, 8]
|
163
|
-
@doubler = lambda{|e| e * 2}
|
164
|
-
end
|
165
|
-
|
166
|
-
it "should only replace columns that actually exist" do
|
167
|
-
lambda{@df.replace!(:not_a_column, &@doubler)}.should raise_error(
|
168
|
-
ArgumentError, /Must provide the name of an existing column./)
|
169
|
-
lambda{@df.replace!(:these, &@doubler)}.should_not raise_error
|
170
|
-
end
|
171
|
-
|
172
|
-
it "should be able to replace a column with a block" do
|
173
|
-
@df.replace!(:these) {|e| e * 2}
|
174
|
-
@df.these.should eql([2,10])
|
175
|
-
end
|
176
|
-
|
177
|
-
it "should be able to replace a column with an array" do
|
178
|
-
@a = [5,9]
|
179
|
-
@df.replace!(:these, @a)
|
180
|
-
@df.these.should eql(@a)
|
181
|
-
end
|
182
|
-
end
|
183
|
-
|
184
|
-
context "filter!" do
|
185
|
-
before do
|
186
|
-
@df.add [1,2,3,4]
|
187
|
-
@df.add [5, 6, 7, 8]
|
188
|
-
end
|
189
|
-
|
190
|
-
it "should be able to filter a data frame with a block using an OpenStruct for each row" do
|
191
|
-
@df.filter!(:open_struct) {|row| row.these == 5}
|
192
|
-
@df.items.should eql([[5, 6, 7, 8]])
|
193
|
-
end
|
194
|
-
|
195
|
-
it "should be able to filter a data frame with a block using a Hash for each row" do
|
196
|
-
@df.filter!(:hash) {|row| row[:these] == 5}
|
197
|
-
@df.items.should eql([[5, 6, 7, 8]])
|
198
|
-
end
|
199
|
-
|
200
|
-
S4 = Struct.new(:one, :two, :three, :four)
|
201
|
-
it "should be able to filter a data frame with a block using another class that uses the row as input" do
|
202
|
-
@df.filter!(S4) {|row| row.one == 5}
|
203
|
-
@df.items.should eql([[5, 6, 7, 8]])
|
204
|
-
end
|
205
|
-
|
206
|
-
it "should be able to filter a data frame with a block using an array for each row" do
|
207
|
-
@df.filter! {|row| row.first == 5}
|
208
|
-
@df.items.should eql([[5, 6, 7, 8]])
|
209
|
-
end
|
210
|
-
|
211
|
-
it "should be able to do fancy things with the row as the filter" do
|
212
|
-
@df.filter! {|row| row.sum > 10}
|
213
|
-
@df.items.should eql([[5, 6, 7, 8]])
|
214
|
-
end
|
215
|
-
|
216
|
-
it "should be able to generate a new data frame with filter" do
|
217
|
-
new_df = @df.filter(:open_struct) {|row| row.these == 5}
|
218
|
-
new_df.items.should eql([[5, 6, 7, 8]])
|
219
|
-
@df.items.should eql([[1, 2, 3, 4], [5, 6, 7, 8]])
|
220
|
-
end
|
221
|
-
|
222
|
-
end
|
223
|
-
|
224
|
-
context "filter_by_category" do
|
225
|
-
|
226
|
-
before do
|
227
|
-
@df = DataFrame.new(:weather, :date)
|
228
|
-
|
229
|
-
(1..31).each do |i|
|
230
|
-
@df.add [(i % 3 == 1) ? :fair : :good, Date.parse("07/#{i}/2009")]
|
231
|
-
end
|
232
|
-
|
233
|
-
@d1 = Date.parse("07/15/2009")
|
234
|
-
@d2 = Date.parse("07/31/2009")
|
235
|
-
|
236
|
-
end
|
237
|
-
|
238
|
-
it "should be able to filter by category" do
|
239
|
-
filtered = @df.filter_by_category(:weather => :good)
|
240
|
-
filtered.weather.uniq.should eql([:good])
|
241
|
-
@df.weather.uniq.should be_include(:fair)
|
242
|
-
end
|
243
|
-
|
244
|
-
it "should be able to manage ranges for filter values" do
|
245
|
-
filtered = @df.filter_by_category(:date => (@d1..@d2))
|
246
|
-
filtered.date.should_not be_include(Date.parse("07/01/2009"))
|
247
|
-
filtered.date.should_not be_include(Date.parse("07/14/2009"))
|
248
|
-
filtered.date.should be_include(Date.parse("07/15/2009"))
|
249
|
-
filtered.date.should be_include(Date.parse("07/31/2009"))
|
250
|
-
@df.date.should be_include(Date.parse("07/01/2009"))
|
251
|
-
end
|
252
|
-
|
253
|
-
it "should be able to take an array of values to filter with" do
|
254
|
-
filtered = @df.filter_by_category(:date => [@d1, @d2])
|
255
|
-
filtered.date.should_not be_include(Date.parse("07/01/2009"))
|
256
|
-
filtered.date.should be_include(Date.parse("07/15/2009"))
|
257
|
-
filtered.date.should be_include(Date.parse("07/31/2009"))
|
258
|
-
end
|
259
|
-
|
260
|
-
it "should have a destructive version" do
|
261
|
-
@df.filter_by_category!(:date => [@d1, @d2])
|
262
|
-
@df.date.should_not be_include(Date.parse("07/01/2009"))
|
263
|
-
@df.date.should be_include(Date.parse("07/15/2009"))
|
264
|
-
@df.date.should be_include(Date.parse("07/31/2009"))
|
265
|
-
end
|
266
|
-
|
267
|
-
end
|
268
|
-
|
269
|
-
context "subset_from_columns" do
|
270
|
-
before do
|
271
|
-
@df.add [1,2,3,4]
|
272
|
-
@df.add [5, 6, 7, 8]
|
273
|
-
end
|
274
|
-
|
275
|
-
it "should be able to create a subset of columns" do
|
276
|
-
new_data_frame = @df.subset_from_columns(:these, :labels)
|
277
|
-
new_data_frame.should_not eql(@df)
|
278
|
-
new_data_frame.labels.should eql([:these, :labels])
|
279
|
-
new_data_frame.items.should eql([[1,4],[5,8]])
|
280
|
-
new_data_frame.these.should eql([1,5])
|
281
|
-
end
|
282
|
-
end
|
283
|
-
|
284
|
-
it "should be able to j_binary_ize! a column, taking its categories and creating a column for each" do
|
285
|
-
df = DataFrame.new(:observations)
|
286
|
-
df.add [:many]
|
287
|
-
df.add [:fine]
|
288
|
-
df.add [:things]
|
289
|
-
df.add [:are]
|
290
|
-
df.add [:available]
|
291
|
-
df.j_binary_ize!(:observations)
|
292
|
-
df.observations_many.should eql([true, false, false, false, false])
|
293
|
-
df.observations_fine.should eql([false, true, false, false, false])
|
294
|
-
df.observations_things.should eql([false, false, true, false, false])
|
295
|
-
df.observations_are.should eql([false, false, false, true, false])
|
296
|
-
df.observations_available.should eql([false, false, false, false, true])
|
297
|
-
df.observations.should eql([:many, :fine, :things, :are, :available])
|
298
|
-
end
|
299
|
-
|
300
|
-
it "should be able to j_binary_ize! a more normal column" do
|
301
|
-
df = DataFrame.new(:observations)
|
302
|
-
df.import([1,2,3,4,5,4,3,2,1].map{|e| Array(e)})
|
303
|
-
df.observations.add_category(:small) {|e| e <= 3}
|
304
|
-
df.observations.add_category(:large) {|e| e >= 3}
|
305
|
-
df.j_binary_ize!(:observations)
|
306
|
-
df.observations_small.should eql([true, true, true, false, false, false, true, true, true])
|
307
|
-
df.observations_large.should eql([false, false, false, true, true, true, false, false, false])
|
308
|
-
end
|
309
140
|
|
310
|
-
it "should be able to j_binary_ize with non-adjacent sets (sets that allow a value to have more than one category)" do
|
311
|
-
df = DataFrame.new(:observations)
|
312
|
-
df.import([1,2,3,4,5,4,3,2,1].map{|e| Array(e)})
|
313
|
-
df.observations.add_category(:small) {|e| e <= 3}
|
314
|
-
df.observations.add_category(:large) {|e| e >= 3}
|
315
|
-
df.j_binary_ize!(:observations, :allow_overlap => true)
|
316
|
-
df.observations_small.should eql([true, true, true, false, false, false, true, true, true])
|
317
|
-
df.observations_large.should eql([false, false, true, true, true, true, true, false, false])
|
318
|
-
end
|
319
|
-
|
320
|
-
it "should be able to hold multiple ideas of a columns categories by resetting the category and re-running j_binary_ize" do
|
321
|
-
df = DataFrame.new(:observations)
|
322
|
-
df.import([1,2,3,4,5,4,3,2,1].map{|e| Array(e)})
|
323
|
-
df.observations.add_category(:small) {|e| e <= 3}
|
324
|
-
df.observations.add_category(:large) {|e| e >= 3}
|
325
|
-
df.j_binary_ize!(:observations, :allow_overlap => true)
|
326
|
-
df.observations.set_categories(:odd => lambda{|e| e.odd?}, :even => lambda{|e| e.even?})
|
327
|
-
df.j_binary_ize!(:observations)
|
328
|
-
df.observations_small.should eql([true, true, true, false, false, false, true, true, true])
|
329
|
-
df.observations_large.should eql([false, false, true, true, true, true, true, false, false])
|
330
|
-
df.observations.should eql([1,2,3,4,5,4,3,2,1])
|
331
|
-
df.observations_even.should eql([false, true, false, true, false, true, false, true, false])
|
332
|
-
df.observations_odd.should eql([true, false, true, false, true, false, true, false, true])
|
333
|
-
end
|
334
|
-
|
335
|
-
context "append!" do
|
336
|
-
|
337
|
-
before do
|
338
|
-
@df.add [1,2,3,4]
|
339
|
-
@df.add [5, 6, 7, 8]
|
340
|
-
end
|
341
|
-
|
342
|
-
it "should be able to append an array of values to the data frame" do
|
343
|
-
@df.append!(:new_column, [5,5])
|
344
|
-
@df.new_column.should eql([5,5])
|
345
|
-
end
|
346
|
-
|
347
|
-
it "should be able to append a default value to the data frame" do
|
348
|
-
@df.append!(:new_column, :value)
|
349
|
-
@df.new_column.should eql([:value, :value])
|
350
|
-
end
|
351
|
-
|
352
|
-
it "should use nil as the default value" do
|
353
|
-
@df.append!(:new_column)
|
354
|
-
@df.new_column.should eql([nil, nil])
|
355
|
-
end
|
356
|
-
end
|
357
141
|
end
|
@@ -0,0 +1,22 @@
|
|
1
|
+
require File.join(File.dirname(__FILE__), "/../spec_helper")
|
2
|
+
require 'data_frame/id3'
|
3
|
+
|
4
|
+
describe "DecisionTree" do
|
5
|
+
|
6
|
+
before do
|
7
|
+
@filename = File.expand_path(File.join(File.dirname(__FILE__), "../fixtures/discrete_training.csv"))
|
8
|
+
@df = DataFrame.from_csv(@filename)
|
9
|
+
@test_data = File.read(@filename)
|
10
|
+
end
|
11
|
+
|
12
|
+
it "should require the decisiontree gem" do
|
13
|
+
defined?(DecisionTree::ID3Tree).should eql('constant')
|
14
|
+
end
|
15
|
+
|
16
|
+
it "should be able to create a decision tree from a data frame" do
|
17
|
+
# Come back to this.
|
18
|
+
# @df.create_id3(:purchase)
|
19
|
+
# @df.id3.train
|
20
|
+
# @df.id3.predict(["36 - 55", "masters", "high", "single", 1]).should eql(1)
|
21
|
+
end
|
22
|
+
end
|
@@ -0,0 +1,13 @@
|
|
1
|
+
require File.join(File.dirname(__FILE__), "/../spec_helper")
|
2
|
+
|
3
|
+
describe Array do
|
4
|
+
it "should be able to determine its dimensions" do
|
5
|
+
[1,2,3].dimensions.should eql(1)
|
6
|
+
[[1,2,3], [1,2,3]].dimensions.should eql(2)
|
7
|
+
[[[1,2,3], [1,2,3]], [[1,2,3], [1,2,3], [[1,2,3], [1,2,3]]]].dimensions.should eql(3)
|
8
|
+
end
|
9
|
+
|
10
|
+
it "should depend on the first element to determine dimensions" do
|
11
|
+
[1, [1,2]].dimensions.should eql(1)
|
12
|
+
end
|
13
|
+
end
|
@@ -0,0 +1,21 @@
|
|
1
|
+
Age,Education,Income,Marital Status,Purchase
|
2
|
+
36 - 55,masters,high,single,will buy
|
3
|
+
18 - 35,high school,low,single,won't buy
|
4
|
+
36 - 55,masters,low,single,will buy
|
5
|
+
18 - 35,bachelors,high,single,won't buy
|
6
|
+
< 18,high school,low,single,will buy
|
7
|
+
18 - 35,bachelors,high,married,won't buy
|
8
|
+
36 - 55,bachelors,low,married,won't buy
|
9
|
+
> 55,bachelors,high,single,will buy
|
10
|
+
36 - 55,masters,low,married,won't buy
|
11
|
+
> 55,masters,low,married,will buy
|
12
|
+
36 - 55,masters,high,single,will buy
|
13
|
+
> 55,masters,high,single,will buy
|
14
|
+
< 18,high school,high,single,won't buy
|
15
|
+
36 - 55,masters,low,single,will buy
|
16
|
+
36 - 55,high school,low,single,will buy
|
17
|
+
< 18,high school,low,married,will buy
|
18
|
+
18 - 35,bachelors,high,married,won't buy
|
19
|
+
> 55,high school,high,married,will buy
|
20
|
+
> 55,bachelors,low,single,will buy
|
21
|
+
36 - 55,high school,high,married,won't buy
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: davidrichards-data_frame
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.0.
|
4
|
+
version: 0.0.19
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- David Richards
|
@@ -9,8 +9,8 @@ autorequire:
|
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
11
|
|
12
|
-
date: 2009-09-
|
13
|
-
default_executable:
|
12
|
+
date: 2009-09-24 00:00:00 -07:00
|
13
|
+
default_executable: plain_frame
|
14
14
|
dependencies:
|
15
15
|
- !ruby/object:Gem::Dependency
|
16
16
|
name: activesupport
|
@@ -44,8 +44,8 @@ dependencies:
|
|
44
44
|
version:
|
45
45
|
description: Data Frames with memoized transpose
|
46
46
|
email: davidlamontrichards@gmail.com
|
47
|
-
executables:
|
48
|
-
|
47
|
+
executables:
|
48
|
+
- plain_frame
|
49
49
|
extensions: []
|
50
50
|
|
51
51
|
extra_rdoc_files: []
|
@@ -53,31 +53,58 @@ extra_rdoc_files: []
|
|
53
53
|
files:
|
54
54
|
- README.rdoc
|
55
55
|
- VERSION.yml
|
56
|
+
- bin/plain_frame
|
56
57
|
- lib/data_frame
|
57
58
|
- lib/data_frame/arff.rb
|
58
59
|
- lib/data_frame/callback_array.rb
|
60
|
+
- lib/data_frame/core
|
61
|
+
- lib/data_frame/core/column_management.rb
|
62
|
+
- lib/data_frame/core/filter.rb
|
63
|
+
- lib/data_frame/core/import.rb
|
64
|
+
- lib/data_frame/core/pre_process.rb
|
65
|
+
- lib/data_frame/core/saving.rb
|
66
|
+
- lib/data_frame/core/training.rb
|
59
67
|
- lib/data_frame/data_frame.rb
|
68
|
+
- lib/data_frame/id3.rb
|
69
|
+
- lib/data_frame/kmeans.rb
|
70
|
+
- lib/data_frame/labels_from_uci.rb
|
71
|
+
- lib/data_frame/mlp.rb
|
60
72
|
- lib/data_frame/model.rb
|
61
73
|
- lib/data_frame/parameter_capture.rb
|
74
|
+
- lib/data_frame/sbn.rb
|
62
75
|
- lib/data_frame/transposable_array.rb
|
63
76
|
- lib/data_frame.rb
|
64
77
|
- lib/ext
|
78
|
+
- lib/ext/array.rb
|
65
79
|
- lib/ext/open_struct.rb
|
66
80
|
- lib/ext/string.rb
|
67
81
|
- lib/ext/symbol.rb
|
68
82
|
- spec/data_frame
|
69
83
|
- spec/data_frame/arff_spec.rb
|
70
84
|
- spec/data_frame/callback_array_spec.rb
|
85
|
+
- spec/data_frame/core
|
86
|
+
- spec/data_frame/core/column_management_spec.rb
|
87
|
+
- spec/data_frame/core/filter_spec.rb
|
88
|
+
- spec/data_frame/core/import_spec.rb
|
89
|
+
- spec/data_frame/core/pre_process_spec.rb
|
90
|
+
- spec/data_frame/core/saving_spec.rb
|
91
|
+
- spec/data_frame/core/training_spec.rb
|
71
92
|
- spec/data_frame/data_frame_spec.rb
|
93
|
+
- spec/data_frame/id3_spec.rb
|
72
94
|
- spec/data_frame/model_spec.rb
|
73
95
|
- spec/data_frame/parameter_capture_spec.rb
|
74
96
|
- spec/data_frame/transposable_array_spec.rb
|
75
97
|
- spec/data_frame_spec.rb
|
98
|
+
- spec/ext
|
99
|
+
- spec/ext/array_spec.rb
|
76
100
|
- spec/fixtures
|
77
101
|
- spec/fixtures/basic.csv
|
102
|
+
- spec/fixtures/discrete_testing.csv
|
103
|
+
- spec/fixtures/discrete_training.csv
|
78
104
|
- spec/spec_helper.rb
|
79
105
|
has_rdoc: true
|
80
106
|
homepage: http://github.com/davidrichards/data_frame
|
107
|
+
licenses:
|
81
108
|
post_install_message:
|
82
109
|
rdoc_options:
|
83
110
|
- --inline-source
|
@@ -99,7 +126,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
99
126
|
requirements: []
|
100
127
|
|
101
128
|
rubyforge_project:
|
102
|
-
rubygems_version: 1.
|
129
|
+
rubygems_version: 1.3.5
|
103
130
|
signing_key:
|
104
131
|
specification_version: 2
|
105
132
|
summary: Data Frames with memoized transpose
|