stat_c 0.1.0 → 0.2.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/README.md +48 -4
- data/benchmark/benchmark.rb +25 -7
- data/ext/stat_c/stat_c.c +178 -30
- data/lib/stat_c/version.rb +2 -1
- data/stat_c.gemspec +2 -2
- metadata +3 -3
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA1:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: e0e5819680b725e08009314b625cd71e806fd2bc
|
4
|
+
data.tar.gz: 707e936d379f538b9ac30661c3d7b8aa99a8d0ab
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 4a2338631efc5343c4181147bf65aa90e0306fb679772c676bbeeada9de57998f24467a012fe2d39cfde2ec1bc6e58ba2b419aa8adf853fd932dbbe7521bbd77
|
7
|
+
data.tar.gz: 70702b3acb197b3b35d6de33d733c00ac86ee11891f6994bb64f8751c0ffa0545f652b6e1e58f50a794848020d3c7358038a4afe7d9d48693cdf59226b3e28b0
|
data/README.md
CHANGED
@@ -1,8 +1,6 @@
|
|
1
1
|
# StatC
|
2
2
|
|
3
|
-
|
4
|
-
|
5
|
-
TODO: Delete this and the text above, and describe your gem
|
3
|
+
Fast, well documented C stats extension for Ruby.
|
6
4
|
|
7
5
|
## Installation
|
8
6
|
|
@@ -22,7 +20,53 @@ Or install it yourself as:
|
|
22
20
|
|
23
21
|
## Usage
|
24
22
|
|
25
|
-
|
23
|
+
```ruby
|
24
|
+
require "stat_c"
|
25
|
+
|
26
|
+
ary = [-1.4, 0, 1, 2, 3.0]
|
27
|
+
|
28
|
+
StatC::Array.mean(ary).round(2) #=> 0.92
|
29
|
+
|
30
|
+
# Stats based on sample variance
|
31
|
+
StatC::Array.var(ary).round(2) #=> 2.93
|
32
|
+
StatC::Array.sd(ary).round(2) #=> 1.71
|
33
|
+
StatC::Array.se(ary).round(2) #=> 0.77
|
34
|
+
|
35
|
+
# Stats based on population variance
|
36
|
+
StatC::Array.var(ary, pop=true).round(2) #=> 2.35
|
37
|
+
StatC::Array.sd(ary, pop=true).round(2) #=> 1.53
|
38
|
+
StatC::Array.se(ary, pop=true).round(2) #=> 0.68
|
39
|
+
```
|
40
|
+
|
41
|
+
## Benchmark ##
|
42
|
+
|
43
|
+
StatC is faster than pure Ruby (duh, it's a C extension {^_^} ). See `benchmark/benchmark.rb` for more info.
|
44
|
+
|
45
|
+
$ ruby benchmark/benchmark.rb
|
46
|
+
|
47
|
+
Rehearsal ----------------------------------------------
|
48
|
+
Ruby mean 0.090000 0.000000 0.090000 ( 0.085029)
|
49
|
+
StatC mean 0.010000 0.000000 0.010000 ( 0.009604)
|
50
|
+
Ruby var 0.350000 0.010000 0.360000 ( 0.357243)
|
51
|
+
StatC var 0.020000 0.000000 0.020000 ( 0.020343)
|
52
|
+
Ruby sd 0.350000 0.000000 0.350000 ( 0.355273)
|
53
|
+
StatC sd 0.020000 0.000000 0.020000 ( 0.018590)
|
54
|
+
Ruby se 0.340000 0.000000 0.340000 ( 0.353170)
|
55
|
+
StatC se 0.030000 0.000000 0.030000 ( 0.025813)
|
56
|
+
------------------------------------- total: 1.220000sec
|
57
|
+
|
58
|
+
user system total real
|
59
|
+
Ruby mean 0.080000 0.000000 0.080000 ( 0.079849)
|
60
|
+
StatC mean 0.000000 0.000000 0.000000 ( 0.009006)
|
61
|
+
|
62
|
+
Ruby var 0.320000 0.010000 0.330000 ( 0.322538)
|
63
|
+
StatC var 0.020000 0.000000 0.020000 ( 0.018962)
|
64
|
+
|
65
|
+
Ruby sd 0.330000 0.000000 0.330000 ( 0.329038)
|
66
|
+
StatC sd 0.020000 0.000000 0.020000 ( 0.020783)
|
67
|
+
|
68
|
+
Ruby se 0.310000 0.000000 0.310000 ( 0.319696)
|
69
|
+
StatC se 0.020000 0.000000 0.020000 ( 0.019259)
|
26
70
|
|
27
71
|
## Development
|
28
72
|
|
data/benchmark/benchmark.rb
CHANGED
@@ -1,18 +1,36 @@
|
|
1
1
|
require "stat_c"
|
2
2
|
require "benchmark"
|
3
3
|
|
4
|
-
def
|
4
|
+
def ary_mean ary
|
5
5
|
ary.reduce(:+) / ary.length.to_f
|
6
6
|
end
|
7
7
|
|
8
|
+
# sample variance
|
9
|
+
def ary_var ary
|
10
|
+
mean = ary_mean ary
|
11
|
+
ary.map { |num| (num - mean) ** 2 }.reduce(:+) / (ary.length - 1)
|
12
|
+
end
|
13
|
+
|
14
|
+
def ary_sd ary
|
15
|
+
Math.sqrt(ary_var ary)
|
16
|
+
end
|
17
|
+
|
18
|
+
def ary_se ary
|
19
|
+
ary_sd(ary) / Math.sqrt(ary.length)
|
20
|
+
end
|
21
|
+
|
8
22
|
ary = (1..1_000_000).map(&:itself)
|
9
23
|
|
10
24
|
Benchmark.bmbm do |x|
|
11
|
-
x.report("
|
12
|
-
|
13
|
-
|
25
|
+
x.report("Ruby mean") { ary_mean ary }
|
26
|
+
x.report("StatC mean") { StatC::Array.mean ary }
|
27
|
+
|
28
|
+
x.report("Ruby var") { ary_var ary }
|
29
|
+
x.report("StatC var") { StatC::Array.var ary }
|
30
|
+
|
31
|
+
x.report("Ruby sd") { ary_sd ary }
|
32
|
+
x.report("StatC sd") { StatC::Array.sd ary }
|
14
33
|
|
15
|
-
x.report("
|
16
|
-
|
17
|
-
end
|
34
|
+
x.report("Ruby se") { ary_se ary }
|
35
|
+
x.report("StatC se") { StatC::Array.se ary }
|
18
36
|
end
|
data/ext/stat_c/stat_c.c
CHANGED
@@ -21,6 +21,10 @@ along with StatC. If not, see <http://www.gnu.org/licenses/>.
|
|
21
21
|
*********************************************************************/
|
22
22
|
|
23
23
|
#include <ruby.h>
|
24
|
+
#include <math.h>
|
25
|
+
|
26
|
+
/* based on NIL_P in ruby.h */
|
27
|
+
#define FALSE_P(v) !((VALUE)(v) != Qfalse)
|
24
28
|
|
25
29
|
/* classes and modules */
|
26
30
|
VALUE sc_mStatC;
|
@@ -28,16 +32,42 @@ VALUE sc_mArray;
|
|
28
32
|
VALUE sc_mError;
|
29
33
|
VALUE sc_eError;
|
30
34
|
|
31
|
-
|
35
|
+
/* @private */
|
36
|
+
static size_t assert_array_not_empty(VALUE ary)
|
32
37
|
{
|
33
|
-
|
34
|
-
unsigned long i = 0;
|
35
|
-
long double sum = 0;
|
36
38
|
size_t len = RARRAY_LEN(ary);
|
37
39
|
|
38
40
|
if (len <= 0) {
|
39
41
|
rb_raise(sc_eError, "Array cannot be empty");
|
42
|
+
} else {
|
43
|
+
return len;
|
40
44
|
}
|
45
|
+
}
|
46
|
+
|
47
|
+
/* @private */
|
48
|
+
static long double sc_ary_entry(VALUE ary, long idx)
|
49
|
+
{
|
50
|
+
return NUM2DBL(rb_ary_entry(ary, idx));
|
51
|
+
}
|
52
|
+
|
53
|
+
/* Calculate the mean of values in the given array.
|
54
|
+
|
55
|
+
@param ary [Array<Numeric>] an array of Numerics
|
56
|
+
|
57
|
+
@example Get mean of array
|
58
|
+
StatC::Array.mean([-1.4, 0, 1, 2, 3.0]).round(2) #=> 0.92
|
59
|
+
|
60
|
+
@raise [StatC::Error::Error] if array length is zero
|
61
|
+
|
62
|
+
@return [Numeric] mean of values in the array
|
63
|
+
|
64
|
+
*/
|
65
|
+
static VALUE sc_mean(VALUE obj, VALUE ary)
|
66
|
+
{
|
67
|
+
|
68
|
+
unsigned long i = 0;
|
69
|
+
long double sum = 0;
|
70
|
+
size_t len = assert_array_not_empty(ary);
|
41
71
|
|
42
72
|
for (i = 0; i < len; ++i) {
|
43
73
|
sum += NUM2DBL(rb_ary_entry(ary, i));
|
@@ -46,44 +76,162 @@ static VALUE sc_mean(VALUE self, VALUE ary)
|
|
46
76
|
return DBL2NUM(sum / len);
|
47
77
|
}
|
48
78
|
|
49
|
-
/*
|
50
|
-
/* { */
|
79
|
+
/* Calculate the variance of values in given array.
|
51
80
|
|
52
|
-
|
81
|
+
If pop param is set to true, calculates the population variance of
|
82
|
+
values in the array. Otherwise, the sample variance is calculated
|
83
|
+
(default).
|
53
84
|
|
54
|
-
|
55
|
-
|
56
|
-
|
85
|
+
@param ary [Array<Numeric>] an array of Numerics
|
86
|
+
@param pop [Bool] pass true to calculate population variance,
|
87
|
+
default: false
|
57
88
|
|
58
|
-
|
59
|
-
|
89
|
+
@example Get sample variance of array
|
90
|
+
StatC::Array.var([-1.4, 0, 1, 2, 3.0]).round(2) #=> 2.93
|
91
|
+
@example Get population variance of array
|
92
|
+
StatC::Array.var([-1.4, 0, 1, 2, 3.0], pop=true).round(2) #=> 2.35
|
60
93
|
|
61
|
-
|
62
|
-
/* if (NIL_P(sample)) { sample = 1; } */
|
63
|
-
/* } */
|
94
|
+
@raise [StatC::Error::Error] if array length is zero
|
64
95
|
|
65
|
-
|
66
|
-
/* { */
|
67
|
-
/* VALUE ary, sample; */
|
96
|
+
@return [Numeric] variance of values in the array
|
68
97
|
|
69
|
-
|
70
|
-
|
98
|
+
*/
|
99
|
+
static VALUE sc_var(int argc, VALUE* argv, VALUE obj)
|
100
|
+
{
|
101
|
+
VALUE ary, calc_pop_var;
|
71
102
|
|
72
|
-
/*
|
73
|
-
|
74
|
-
/* } */
|
103
|
+
/* one required and one optional argument */
|
104
|
+
rb_scan_args(argc, argv, "11", &ary, &calc_pop_var);
|
75
105
|
|
76
|
-
|
106
|
+
unsigned long i = 0;
|
107
|
+
long double sum = 0;
|
108
|
+
size_t len = assert_array_not_empty(ary);
|
109
|
+
|
110
|
+
long double mean = NUM2DBL(sc_mean(obj, ary));
|
111
|
+
|
112
|
+
for (i = 0; i < len; ++i) {
|
113
|
+
sum += pow(sc_ary_entry(ary, i) - mean, 2);
|
114
|
+
}
|
115
|
+
|
116
|
+
if (NIL_P(calc_pop_var) || FALSE_P(calc_pop_var)) { /* sample variance */
|
117
|
+
return DBL2NUM(sum / (len - 1));
|
118
|
+
} else { /* population variance */
|
119
|
+
return DBL2NUM(sum / len);
|
120
|
+
}
|
121
|
+
}
|
122
|
+
|
123
|
+
/* Calculate the standard deviation of values in given array.
|
124
|
+
|
125
|
+
If pop param is set to true, the standard deviation is based on
|
126
|
+
population variance. Otherwise, sample variance is used (default).
|
127
|
+
|
128
|
+
@param ary [Array<Numeric>] an array of Numerics
|
129
|
+
@param pop [Bool] pass true to calculate population standard
|
130
|
+
deviation, default: false
|
131
|
+
|
132
|
+
@example Get sample standard deviation of array
|
133
|
+
StatC::Array.sd([-1.4, 0, 1, 2, 3.0]).round(2) #=> 1.71
|
134
|
+
@example Get population standard deviation of array
|
135
|
+
StatC::Array.sd([-1.4, 0, 1, 2, 3.0], pop=true).round(2) #=> 1.53
|
136
|
+
|
137
|
+
@raise [StatC::Error::Error] if array length is zero
|
138
|
+
|
139
|
+
@return [Numeric] standard deviation of values in the array
|
140
|
+
|
141
|
+
*/
|
142
|
+
static VALUE sc_sd(int argc, VALUE* argv, VALUE obj)
|
77
143
|
{
|
78
|
-
|
144
|
+
VALUE ary, calc_pop_var;
|
79
145
|
|
146
|
+
/* one required and one optional argument */
|
147
|
+
rb_scan_args(argc, argv, "11", &ary, &calc_pop_var);
|
148
|
+
|
149
|
+
return DBL2NUM(sqrt(NUM2DBL(sc_var(argc, argv, obj))));
|
150
|
+
}
|
151
|
+
|
152
|
+
/* Calculate the standard deviation of values in given array.
|
153
|
+
|
154
|
+
If pop param is set to true, the standard error of the mean is based
|
155
|
+
on population variance. Otherwise, sample variance is used (default).
|
156
|
+
|
157
|
+
@param ary [Array<Numeric>] an array of Numerics
|
158
|
+
@param pop [Bool] pass true to calculate population standard
|
159
|
+
error of the mean, default: false
|
160
|
+
|
161
|
+
@example Get sample standard error of array
|
162
|
+
StatC::Array.se([-1.4, 0, 1, 2, 3.0]).round(2) #=> 0.77
|
163
|
+
@example Get population standard error of array
|
164
|
+
StatC::Array.se([-1.4, 0, 1, 2, 3.0], pop=true).round(2) #=> 0.68
|
165
|
+
|
166
|
+
@raise [StatC::Error::Error] if array length is zero
|
167
|
+
|
168
|
+
@return [Numeric] standard error of the mean for values in the array
|
169
|
+
|
170
|
+
*/
|
171
|
+
static VALUE sc_se(int argc, VALUE* argv, VALUE obj)
|
172
|
+
{
|
173
|
+
VALUE ary, calc_pop_var;
|
174
|
+
|
175
|
+
/* one required and one optional argument */
|
176
|
+
rb_scan_args(argc, argv, "11", &ary, &calc_pop_var);
|
177
|
+
|
178
|
+
long double sd = NUM2DBL(sc_sd(argc, argv, obj));
|
179
|
+
|
180
|
+
size_t len = assert_array_not_empty(ary);
|
181
|
+
|
182
|
+
return DBL2NUM(sd / sqrt(len));
|
183
|
+
}
|
184
|
+
|
185
|
+
/* Document-module: StatC::Array
|
186
|
+
|
187
|
+
Statistical methods operating on the values of an array
|
188
|
+
|
189
|
+
*/
|
190
|
+
static void sc_init_mArray(void)
|
191
|
+
{
|
80
192
|
sc_mArray = rb_define_module_under(sc_mStatC, "Array");
|
193
|
+
|
194
|
+
rb_define_singleton_method(sc_mArray, "mean", sc_mean, 1);
|
195
|
+
rb_define_singleton_method(sc_mArray, "var", sc_var, -1);
|
196
|
+
rb_define_singleton_method(sc_mArray, "sd", sc_sd, -1);
|
197
|
+
rb_define_singleton_method(sc_mArray, "se", sc_se, -1);
|
198
|
+
}
|
199
|
+
|
200
|
+
/* Document-module: StatC::Error
|
201
|
+
|
202
|
+
Module containing all error classes of the StatC module.
|
203
|
+
|
204
|
+
*/
|
205
|
+
static void sc_init_mError(void)
|
206
|
+
{
|
81
207
|
sc_mError = rb_define_module_under(sc_mStatC, "Error");
|
208
|
+
}
|
82
209
|
|
83
|
-
sc_eError = rb_define_class_under(sc_mError, "Error", rb_eStandardError);
|
84
210
|
|
85
|
-
|
86
|
-
|
87
|
-
|
88
|
-
|
211
|
+
/* Document-class: StatC::Error::Error
|
212
|
+
|
213
|
+
Error class from which all errors raised by StatC inherit. Thus, you
|
214
|
+
can rescue from StatC::Error::Error to catch all errors specific to
|
215
|
+
StatC.
|
216
|
+
|
217
|
+
@note Inherits from StandardError
|
218
|
+
|
219
|
+
*/
|
220
|
+
static void sc_init_eError(void)
|
221
|
+
{
|
222
|
+
sc_eError = rb_define_class_under(sc_mError, "Error", rb_eStandardError);
|
223
|
+
}
|
224
|
+
|
225
|
+
/* Document-module: StatC
|
226
|
+
|
227
|
+
C stats module for Ruby.
|
228
|
+
|
229
|
+
*/
|
230
|
+
void Init_stat_c(void)
|
231
|
+
{
|
232
|
+
sc_mStatC = rb_define_module("StatC");
|
233
|
+
|
234
|
+
sc_init_mArray();
|
235
|
+
sc_init_mError();
|
236
|
+
sc_init_eError();
|
89
237
|
}
|
data/lib/stat_c/version.rb
CHANGED
data/stat_c.gemspec
CHANGED
@@ -10,8 +10,8 @@ Gem::Specification.new do |spec|
|
|
10
10
|
spec.email = ["moorer@udel.edu"]
|
11
11
|
spec.license = "GPLv3"
|
12
12
|
|
13
|
-
spec.summary = %q{Fast
|
14
|
-
spec.description = %q{Fast
|
13
|
+
spec.summary = %q{Fast, well documented C stats extension for Ruby.}
|
14
|
+
spec.description = %q{Fast, well documented C stats extension for Ruby.}
|
15
15
|
spec.homepage = "https://github.com/mooreryan/stat_c"
|
16
16
|
|
17
17
|
spec.files = `git ls-files -z`.split("\x0").reject { |f| f.match(%r{^(test|spec|features)/}) }
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: stat_c
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.
|
4
|
+
version: 0.2.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Ryan Moore
|
@@ -114,7 +114,7 @@ dependencies:
|
|
114
114
|
- - "~>"
|
115
115
|
- !ruby/object:Gem::Version
|
116
116
|
version: 0.8.11
|
117
|
-
description: Fast
|
117
|
+
description: Fast, well documented C stats extension for Ruby.
|
118
118
|
email:
|
119
119
|
- moorer@udel.edu
|
120
120
|
executables: []
|
@@ -164,6 +164,6 @@ rubyforge_project:
|
|
164
164
|
rubygems_version: 2.5.1
|
165
165
|
signing_key:
|
166
166
|
specification_version: 4
|
167
|
-
summary: Fast
|
167
|
+
summary: Fast, well documented C stats extension for Ruby.
|
168
168
|
test_files: []
|
169
169
|
has_rdoc:
|