1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
|
require 'rubygems'
require 'opentox-ruby'
require 'test/unit'
require 'validate-owl'
require 'test-util'
DELETE = true
class RUtilTest < Test::Unit::TestCase
include TestUtil
def global_setup
unless defined?(@@subjectid)
@@subjectid = OpenTox::Authorization.authenticate("guest","guest")
@@signout = true
end
@@rutil = OpenTox::RUtil.new
@@hamster = OpenTox::Dataset.create_from_csv_file(File.new("data/hamster_carcinogenicity.csv").path, @@subjectid)
pred_feature = @@hamster.features.keys[0]
fminer = File.join(CONFIG[:services]["opentox-algorithm"],"fminer/bbrc")
hamster_features_uri = OpenTox::RestClientWrapper.post(fminer,
{:dataset_uri=>@@hamster.uri,:prediction_feature=>pred_feature,:subjectid=>@@subjectid}).to_s
@@hamster_features = OpenTox::Dataset.find(hamster_features_uri,@@subjectid)
@@resources = [@@hamster.uri, hamster_features_uri]
end
def global_teardown
@@rutil.quit_r
if DELETE
@@resources.each do |uri|
OpenTox::RestClientWrapper.delete(uri,{:subjectid=>@@subjectid})
end
else
puts "Resources "+@@resources.to_yaml
end
OpenTox::Authorization.logout(@@subjectid) if defined?(@@signout)
end
def pre_files(files)
files.each{|f| File.delete(f) if File.exist?(f)}
end
def post_files(files)
files.each{|f| assert File.exist?(f)}
if DELETE
files.each{|f| File.delete(f) if File.exist?(f)}
else
puts "Plotted to "+files.to_yaml
end
end
def test_paired_ttest
puts "ttest"
x = Array.new(1000,0).collect{|e| rand()}
y = Array.new(1000,0).collect{|e| rand()}
res = @@rutil.paired_ttest(x,y)
puts "x >> y ? #{res}"
assert_equal res,0
y = Array.new(1000,0).collect{|e| rand()-0.1}
res = @@rutil.paired_ttest(x,y)
puts "x >> y ? #{res}"
assert res>0
end
def test_boxplot
puts "boxplot"
files = ["/tmp/box.svg","/tmp/box.png"]
pre_files(files)
data = [
[ :method, [4,4,5,5,4,3,2] ],
[ :method2, [1,2,3,4,5,4,6] ],
[ :asdf, [9,1,8,0,7,1,6] ] ]
@@rutil.boxplot(files, data, "comparison1" )
post_files(files)
end
def test_double_hist_plot
puts "double_hist_plot"
hist_num_log = "/tmp/hist_num_log.svg"
hist_num = "/tmp/hist_num.svg"
hist_cat = "/tmp/hist_cat.svg"
pre_files [hist_num_log,hist_num,hist_cat]
data1 = Array.new(1000,0).collect{|x| rand()*rand()}
data2 = Array.new(1000,0).collect{|x| rand()*rand()*rand()}
@@rutil.double_hist_plot([hist_num_log], data1, data2, true, true )
@@rutil.double_hist_plot([hist_num], data1, data2, true, false )
data1 = "a,a,a,a,b,b,b,b,b,b,b,b,b,b,c,c,c".split(",")
data2 = "a,a,a,a,a,a,b,b,b,b,b,b,c,c,c,c,c,d,d,d,d,d".split(",")
@@rutil.double_hist_plot([hist_cat], data1, data2, false )
post_files [hist_num_log,hist_num,hist_cat]
end
def test_dataset_to_dataframe
puts "dataset_to_dataframe"
dataframe = @@rutil.dataset_to_dataframe(@@hamster,"NA",@@subjectid)
dataset_conv = @@rutil.dataframe_to_dataset(dataframe,{},@@subjectid)
dataset_conv_reloaded = OpenTox::Dataset.find(dataset_conv.uri,@@subjectid)
@@resources << dataset_conv.uri
dataset_equal(@@hamster,dataset_conv)
dataset_equal(@@hamster,dataset_conv_reloaded)
feats = @@hamster_features.features.keys[0..(@@hamster_features.features.keys.size/2)]
dataframe = @@rutil.dataset_to_dataframe(@@hamster_features,"NA",@@subjectid,feats)
dataset_conv = @@rutil.dataframe_to_dataset(dataframe,{},@@subjectid)
dataset_conv_reloaded = OpenTox::Dataset.find(dataset_conv.uri,@@subjectid)
@@resources << dataset_conv.uri
[dataset_conv, dataset_conv_reloaded].each do |d|
assert_equal d.compounds.sort,@@hamster_features.compounds.sort
assert_equal d.features.keys.sort,feats.sort
d.compounds.each do |c|
d.features.keys.each do |f|
if @@hamster_features.data_entries[c]==nil || @@hamster_features.data_entries[c][f]==nil
assert d.data_entries[c]==nil || d.data_entries[c][f]==nil
else
assert_not_nil d.data_entries[c]
assert_equal @@hamster_features.data_entries[c][f].collect{|v|v.to_s},d.data_entries[c][f].collect{|v|v.to_s}
end
end
end
end
dataframe = @@rutil.dataset_to_dataframe(@@hamster_features,"NA",@@subjectid)
@@rutil.r.eval "#{dataframe} <- #{dataframe}[2:10,10:20]"
dataset_conv = @@rutil.dataframe_to_dataset(dataframe,{},@@subjectid)
dataset_conv_reloaded = OpenTox::Dataset.find(dataset_conv.uri,@@subjectid)
@@resources << dataset_conv.uri
[dataset_conv,dataset_conv_reloaded].each do |d|
assert_equal d.compounds.size,9
d.compounds.size.times do |i|
assert_equal d.compounds[i],@@hamster_features.compounds[i+1]
end
assert_equal d.features.size,11
d.compounds.each do |c|
d.features.keys.each do |f|
if @@hamster_features.data_entries[c]==nil || @@hamster_features.data_entries[c][f]==nil
assert d.data_entries[c]==nil || d.data_entries[c][f]==nil
else
assert_not_nil d.data_entries[c]
assert_equal @@hamster_features.data_entries[c][f].collect{|v|v.to_s},d.data_entries[c][f].collect{|v|v.to_s}
end
end
end
end
end
def stratified_split
unless defined?@@strat
@@split_ratio = 0.05
@@split_has_duplicates = false #hamster has no duplicates
# res = @@rutil.stratified_split(@@hamster,0,@@split_ratio,1)
# @@resources += [ res[0].uri, res[1].uri ]
# @@strat = { :data => @@hamster, :split1 => res[0], :split2 => res[1] }
data_combined = OpenTox::Dataset.merge(@@hamster,@@hamster_features,{},@@subjectid)
res1, res2 = @@rutil.stratified_split(data_combined,{},0,@@split_ratio,@@subjectid,1)
@@resources += [ data_combined.uri, res1.uri, res2.uri ]
@@strat = {:data => data_combined, :split1 => res1, :split2 => res2 }
end
@@strat
end
def test_stratified_split
puts "test_stratified_split"
split = stratified_split
size = split[:data].compounds.size
size1 = split[:split1].compounds.size
size2 = split[:split2].compounds.size
assert_equal size,(split[:split1].compounds+split[:split2].compounds).uniq.size
unless @@split_has_duplicates
assert_equal (@@split_ratio*size).round,size1,
"Dataset #{size} should be split into #{(@@split_ratio*size).round}/#{size-(@@split_ratio*size).round}"+
" (exact: #{@@split_ratio*size}), instead: #{size1}/#{size2}"
end
split[:data].compounds.each do |c|
include1 = split[:split1].compounds.include?(c)
include2 = split[:split2].compounds.include?(c)
unless @@split_has_duplicates
assert(((include1 and !include2) or (!include1 and include2)))
else
assert((include1 or include2))
end
end
end
def test_k_fold_stratified_split
puts "test_k_fold_stratified_split"
data_combined = OpenTox::Dataset.merge(@@hamster,@@hamster_features,{},@@subjectid)
num_duplicates = 0 #hamster has no duplicates
num_folds = 10
avg_split_size = (data_combined.compounds.size+num_duplicates)/num_folds.to_f
@@resources += [ data_combined.uri ]
train, test = @@rutil.stratified_k_fold_split(data_combined,{},0,num_folds,@@subjectid,1)
@@resources += (train + test).collect{ |r| r.uri }
[train, test].each do |result|
assert result.is_a?(Array)
assert result.size==num_folds
end
sum_test = 0
compounds_test = []
num_folds.times do |i|
assert test[i].is_a?(OpenTox::Dataset)
assert test[i].compounds.size==avg_split_size.to_i ||
test[i].compounds.size==(avg_split_size+1).to_i
sum_test+=test[i].compounds.size
compounds_test += test[i].compounds
assert_equal (test[i].compounds.size+train[i].compounds.size),(data_combined.compounds.size+num_duplicates)
compounds = (test[i].compounds + train[i].compounds)
if num_duplicates==0
assert_equal compounds.sort,data_combined.compounds.sort
else
assert_equal compounds.uniq.sort,data_combined.compounds.sort
end
end
assert_equal sum_test,(data_combined.compounds.size+num_duplicates)
if num_duplicates==0
assert_equal compounds_test.sort,data_combined.compounds.sort
else
assert_equal compounds_test.uniq.sort,data_combined.compounds.sort
end
end
def test_feature_value_plot
puts "feature_value_plot"
split = stratified_split
data = split[:data]
dataset1 = data.split( data.compounds[0..4], data.features.keys, {}, @@subjectid)
dataset2 = data.split( data.compounds[5..-1], data.features.keys, {}, @@subjectid)
@@resources += [dataset1.uri, dataset2.uri]
files = []
#plot
#[true,false].each do |fast_embedding|
# next if fast_embedding==false and !@@rutil.package_installed?("smacof")
random_file = "/tmp/feature_value_plot_random.svg" #_fast#{fast_embedding}.svg"
stratified_file = "/tmp/feature_value_plot_stratified.svg" #_fast#{fast_embedding}.svg"
pre_files [random_file, stratified_file]
@@rutil.feature_value_plot([random_file], dataset1.uri, dataset2.uri,
"first five", "rest", nil, @@subjectid)
@@rutil.feature_value_plot([stratified_file], split[:split1].uri, split[:split2].uri,
"five percent stratified", "rest", nil, @@subjectid)
files += [random_file, stratified_file]
#end
#cleanup
post_files files
end
# hack to have a global_setup and global_teardown
def teardown
if((@@expected_test_count-=1) == 0)
global_teardown
end
end
def setup
unless defined?@@expected_test_count
@@expected_test_count = (self.class.instance_methods.reject{|method| method[0..3] != 'test'}).length
global_setup
end
end
end
|