davidrichards-data_frame 0.0.12 → 0.0.13
Sign up to get free protection for your applications and to get access to all the features.
- data/README.rdoc +27 -0
- data/VERSION.yml +1 -1
- data/lib/data_frame.rb +104 -2
- data/lib/ext/open_struct.rb +5 -0
- data/spec/data_frame_spec.rb +60 -0
- metadata +3 -2
data/README.rdoc
CHANGED
@@ -48,6 +48,33 @@ To get your feet wet, you may want to play with data sets found here:
|
|
48
48
|
|
49
49
|
http://www.liaad.up.pt/~ltorgo/Regression/DataSets.html
|
50
50
|
|
51
|
+
== Transformations
|
52
|
+
|
53
|
+
A lot of the work in the data frame is to transform the actual table. You may need to drop columns, filter results, replace values in a column or create a new data frame based on the existing one. Here's how to do that:
|
54
|
+
|
55
|
+
> df = DataFrame.from_csv('http://archive.ics.uci.edu/ml/machine-learning-databases/forest-fires/forestfires.csv')
|
56
|
+
# => DataFrame rows: 517 labels: [:x, :y, :month, :day, :ffmc, :dmc, :dc, :isi, :temp, :rh, :wind, :rain, :area]
|
57
|
+
> df.drop!(:ffmc)
|
58
|
+
# => DataFrame rows: 517 labels: [:x, :y, :month, :day, :dmc, :dc, :isi, :temp, :rh, :wind, :rain, :area]
|
59
|
+
> df.drop!(:dmc, :dc, :isi, :rh)
|
60
|
+
# => DataFrame rows: 517 labels: [:x, :y, :month, :day, :temp, :wind, :rain, :area]
|
61
|
+
> df.x
|
62
|
+
# => [7, 7, 7, 8, 8, 8, 8, 8, 8, 7, 7, 7, 6, 6, 6,...]
|
63
|
+
> df.replace!(:x) {|e| e * 3}
|
64
|
+
# => DataFrame rows: 517 labels: [:x, :y, :month, :day, :temp, :wind, :rain, :area]
|
65
|
+
> df.x
|
66
|
+
# => [21, 21, 21, 24, 24, 24, 24, 24, 24, 21, 21, 21, 18, 18, 18,...]
|
67
|
+
> df.filter!(:open_struct) {|row| row.x == 24}
|
68
|
+
# => DataFrame rows: 61 labels: [:x, :y, :month, :day, :temp, :wind, :rain, :area]
|
69
|
+
> df.x
|
70
|
+
# => [24, 24, 24, 24, 24, 24, 24, 24, 24,...]
|
71
|
+
> new_data_frame = df.subset_from_columns(:x, :y)
|
72
|
+
# => DataFrame rows: 61 labels: [:x, :y]
|
73
|
+
> new_data_frame.items
|
74
|
+
# => [[24, 6], [24, 6], [24, 6], [24, 6], ...]
|
75
|
+
|
76
|
+
|
77
|
+
Note: most of these transformations are not optimized. I'll work with things for a while before I try to optimize this library. However, I should say that I've used some fairly large data sets (thousands of rows) and have been fine with things so far.
|
51
78
|
|
52
79
|
==Installation
|
53
80
|
|
data/VERSION.yml
CHANGED
data/lib/data_frame.rb
CHANGED
@@ -3,6 +3,7 @@ require 'activesupport'
|
|
3
3
|
require 'just_enumerable_stats'
|
4
4
|
require 'open-uri'
|
5
5
|
require 'fastercsv'
|
6
|
+
require 'ostruct'
|
6
7
|
|
7
8
|
# Use a Dictionary if available
|
8
9
|
begin
|
@@ -57,6 +58,9 @@ class DataFrame
|
|
57
58
|
return nil unless contents
|
58
59
|
table = FCSV.parse(contents, default_csv_opts.merge(opts))
|
59
60
|
labels = table.shift
|
61
|
+
while table.last.empty?
|
62
|
+
table.pop
|
63
|
+
end
|
60
64
|
[labels, table]
|
61
65
|
end
|
62
66
|
|
@@ -71,6 +75,10 @@ class DataFrame
|
|
71
75
|
end
|
72
76
|
end
|
73
77
|
|
78
|
+
def inspect
|
79
|
+
"DataFrame rows: #{self.rows.size} labels: #{self.labels.inspect}"
|
80
|
+
end
|
81
|
+
|
74
82
|
# The labels of the data items
|
75
83
|
attr_reader :labels
|
76
84
|
alias :variables :labels
|
@@ -142,14 +150,108 @@ class DataFrame
|
|
142
150
|
end
|
143
151
|
end
|
144
152
|
|
145
|
-
def drop!(
|
153
|
+
def drop!(*labels)
|
154
|
+
labels.each do |label|
|
155
|
+
drop_one!(label)
|
156
|
+
end
|
157
|
+
self
|
158
|
+
end
|
159
|
+
|
160
|
+
def drop_one!(label)
|
146
161
|
i = self.labels.index(label)
|
147
162
|
return nil unless i
|
148
163
|
self.items.each do |item|
|
149
164
|
item.delete_at(i)
|
150
165
|
end
|
151
166
|
self.labels.delete_at(i)
|
152
|
-
|
167
|
+
self
|
168
|
+
end
|
169
|
+
protected :drop_one!
|
170
|
+
|
171
|
+
def replace!(column, values=nil, &block)
|
172
|
+
column = validate_column(column)
|
173
|
+
if not values
|
174
|
+
values = self.send(column)
|
175
|
+
values.map! {|e| block.call(e)}
|
176
|
+
end
|
177
|
+
replace_column(column, values)
|
178
|
+
self
|
179
|
+
end
|
180
|
+
|
181
|
+
def replace_column(column, values)
|
182
|
+
column = validate_column(column)
|
183
|
+
index = self.labels.index(column)
|
184
|
+
list = []
|
185
|
+
self.items.each_with_index do |item, i|
|
186
|
+
consolidated = item
|
187
|
+
consolidated[index] = values[i]
|
188
|
+
list << consolidated
|
189
|
+
end
|
190
|
+
@items = list.dup
|
191
|
+
end
|
192
|
+
protected :replace_column
|
193
|
+
|
194
|
+
def validate_column(column)
|
195
|
+
column = column.to_sym
|
196
|
+
raise ArgumentError, "Must provide the name of an existing column. Provided #{column.inspect}, needed to provide one of #{self.labels.inspect}" unless self.labels.include?(column)
|
197
|
+
column
|
198
|
+
end
|
199
|
+
protected :validate_column
|
200
|
+
|
201
|
+
# Takes a block to evaluate on each row. The row can be converted into
|
202
|
+
# an OpenStruct or a Hash for easier filter methods. Note, don't try this
|
203
|
+
# with a hash or open struct unless you have facets available.
|
204
|
+
def filter!(as=Array, &block)
|
205
|
+
as = infer_class(as)
|
206
|
+
items = []
|
207
|
+
self.items.each do |row|
|
208
|
+
value = block.call(cast_row(row, as))
|
209
|
+
items << row if value
|
210
|
+
end
|
211
|
+
@items = items.dup
|
212
|
+
self
|
213
|
+
end
|
214
|
+
|
215
|
+
def infer_class(obj)
|
216
|
+
obj = obj.to_s.classify.constantize if obj.is_a?(Symbol)
|
217
|
+
obj = obj.classify.constantize if obj.is_a?(String)
|
218
|
+
obj
|
219
|
+
end
|
220
|
+
protected :infer_class
|
221
|
+
|
222
|
+
def cast_row(row, as)
|
223
|
+
if as == Hash
|
224
|
+
obj = {}
|
225
|
+
self.labels.each_with_index do |label, i|
|
226
|
+
obj[label] = row[i]
|
227
|
+
end
|
228
|
+
obj
|
229
|
+
elsif as == OpenStruct
|
230
|
+
obj = OpenStruct.new
|
231
|
+
self.labels.each_with_index do |label, i|
|
232
|
+
obj.table[label] = row[i]
|
233
|
+
end
|
234
|
+
obj
|
235
|
+
elsif as == Array
|
236
|
+
row
|
237
|
+
else
|
238
|
+
as.new(row)
|
239
|
+
end
|
240
|
+
end
|
241
|
+
protected :cast_row
|
242
|
+
|
243
|
+
# Creates a new data frame, only with the specified columns.
|
244
|
+
def subset_from_columns(*cols)
|
245
|
+
new_labels = self.labels.inject([]) do |list, label|
|
246
|
+
list << label if cols.include?(label)
|
247
|
+
list
|
248
|
+
end
|
249
|
+
new_data_frame = DataFrame.new(*self.labels)
|
250
|
+
new_data_frame.import(self.items)
|
251
|
+
self.labels.each do |label|
|
252
|
+
new_data_frame.drop!(label) unless new_labels.include?(label)
|
253
|
+
end
|
254
|
+
new_data_frame
|
153
255
|
end
|
154
256
|
|
155
257
|
end
|
data/spec/data_frame_spec.rb
CHANGED
@@ -101,6 +101,14 @@ describe DataFrame do
|
|
101
101
|
@df.labels.should eql([:threes, :fours])
|
102
102
|
end
|
103
103
|
|
104
|
+
it "should be able to remove more than one column at a time" do
|
105
|
+
@df = DataFrame.new :twos, :threes, :fours
|
106
|
+
@df.import([[2,3,4], [2,3,4], [2,3,4], [2,3,4]])
|
107
|
+
@df.drop!(:twos, :fours)
|
108
|
+
@df.items.all? {|i| i.should eql([3])}
|
109
|
+
@df.labels.should eql([:threes])
|
110
|
+
end
|
111
|
+
|
104
112
|
it "should offer a hash-like structure of columns" do
|
105
113
|
@df.add [1,2,3,4]
|
106
114
|
@df.add [5, 6, 7, 8]
|
@@ -131,4 +139,56 @@ describe DataFrame do
|
|
131
139
|
@df.labels.should eql(@labels)
|
132
140
|
@df.variables.should eql(@labels)
|
133
141
|
end
|
142
|
+
|
143
|
+
context "replace!" do
|
144
|
+
before do
|
145
|
+
@df.add [1,2,3,4]
|
146
|
+
@df.add [5, 6, 7, 8]
|
147
|
+
@doubler = lambda{|e| e * 2}
|
148
|
+
end
|
149
|
+
|
150
|
+
it "should only replace columns that actually exist" do
|
151
|
+
lambda{@df.replace!(:not_a_column, &@doubler)}.should raise_error(
|
152
|
+
ArgumentError, /Must provide the name of an existing column./)
|
153
|
+
lambda{@df.replace!(:these, &@doubler)}.should_not raise_error
|
154
|
+
end
|
155
|
+
|
156
|
+
it "should be able to replace a column with a block" do
|
157
|
+
@df.replace!(:these) {|e| e * 2}
|
158
|
+
@df.these.should eql([2,10])
|
159
|
+
end
|
160
|
+
|
161
|
+
it "should be able to replace a column with an array" do
|
162
|
+
@a = [5,9]
|
163
|
+
@df.replace!(:these, @a)
|
164
|
+
@df.these.should eql(@a)
|
165
|
+
end
|
166
|
+
end
|
167
|
+
|
168
|
+
context "filter!" do
|
169
|
+
before do
|
170
|
+
@df.add [1,2,3,4]
|
171
|
+
@df.add [5, 6, 7, 8]
|
172
|
+
end
|
173
|
+
|
174
|
+
it "should be able to filter a data frame with a block" do
|
175
|
+
@df.filter!(:open_struct) {|row| row.these == 5}
|
176
|
+
@df.items.should eql([[5, 6, 7, 8]])
|
177
|
+
end
|
178
|
+
end
|
179
|
+
|
180
|
+
context "subset_from_columns" do
|
181
|
+
before do
|
182
|
+
@df.add [1,2,3,4]
|
183
|
+
@df.add [5, 6, 7, 8]
|
184
|
+
end
|
185
|
+
|
186
|
+
it "should be able to create a subset of columns" do
|
187
|
+
new_data_frame = @df.subset_from_columns(:these, :labels)
|
188
|
+
new_data_frame.should_not eql(@df)
|
189
|
+
new_data_frame.labels.should eql([:these, :labels])
|
190
|
+
new_data_frame.items.should eql([[1,4],[5,8]])
|
191
|
+
new_data_frame.these.should eql([1,5])
|
192
|
+
end
|
193
|
+
end
|
134
194
|
end
|
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: 0.0.13
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- David Richards
|
@@ -9,7 +9,7 @@ autorequire:
|
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
11
|
|
12
|
-
date: 2009-08-
|
12
|
+
date: 2009-08-16 00:00:00 -07:00
|
13
13
|
default_executable:
|
14
14
|
dependencies:
|
15
15
|
- !ruby/object:Gem::Dependency
|
@@ -58,6 +58,7 @@ files:
|
|
58
58
|
- lib/data_frame/transposable_array.rb
|
59
59
|
- lib/data_frame.rb
|
60
60
|
- lib/ext
|
61
|
+
- lib/ext/open_struct.rb
|
61
62
|
- lib/ext/string.rb
|
62
63
|
- lib/ext/symbol.rb
|
63
64
|
- spec/data_frame
|