bernoulli 0.1.0 → 0.2.0
Sign up to get free protection for your applications and to get access to all the features.
- data/README.md +46 -6
- data/Rakefile +6 -3
- data/bernoulli.gemspec +8 -7
- data/lib/bernoulli/math.rb +0 -1
- data/lib/bernoulli.rb +25 -4
- data/test/test_bernoulli.rb +3 -21
- data/test/test_math.rb +39 -0
- metadata +6 -4
data/README.md
CHANGED
@@ -1,4 +1,3 @@
|
|
1
|
-
|
2
1
|
# Bernoulli : A Ruby library for the binomial distribution
|
3
2
|
|
4
3
|
In probability theory and statistics, the **binomial distribution** is the discrete probability distribution of the number of successes in a sequence of *n* independent yes/no experiments, each of which yields success with probability *p*.
|
@@ -11,23 +10,23 @@ Use RubyGems to install `bernoulli`
|
|
11
10
|
|
12
11
|
$ gem install bernoulli
|
13
12
|
|
14
|
-
##
|
13
|
+
## Tutorial
|
15
14
|
|
16
15
|
Using `bernoulli`, one can calculate the probabilities of bernoulli experiments.
|
17
16
|
|
18
|
-
|
17
|
+
Let's start by looking at an example.
|
19
18
|
|
20
19
|
**Question:** What is the chance of getting between 25 and 36 sixes when one rolls a die 180 times? What is the expected value and variance of this trial?
|
21
20
|
|
22
21
|
Let's use `bernoulli` to solve this:
|
23
22
|
|
24
|
-
First we load up `irb` from the terminal including `bernoulli` as a library
|
23
|
+
First we load up `irb` from the terminal including `bernoulli` as a library.
|
25
24
|
|
26
25
|
$ irb -r bernoulli
|
27
26
|
|
28
|
-
Or we could just
|
27
|
+
Or we could just `require 'bernoulli'` in any Ruby script.
|
29
28
|
|
30
|
-
Then we create a new instance of class Bernoulli and calculate our desired probability and properties
|
29
|
+
Then we create a new instance of class Bernoulli and calculate our desired probability and properties like so
|
31
30
|
|
32
31
|
x = Bernoulli.new(180, 1.0/6)
|
33
32
|
# => #<Bernoulli:0x007fba721a1bf0 @n=180, @p=0.16666666666666666>
|
@@ -38,3 +37,44 @@ Then we create a new instance of class Bernoulli and calculate our desired proba
|
|
38
37
|
x.expected_value # => 30.0
|
39
38
|
x.variance # => 25.0
|
40
39
|
|
40
|
+
We could also calculate the standard deviation, excess or skewness:
|
41
|
+
|
42
|
+
x.standard_deviation # => 5.0
|
43
|
+
x.excess # => 0.006666666666666665
|
44
|
+
x.skewness => 0.13333333333333336
|
45
|
+
|
46
|
+
`bernoulli` can also do empirical tests. Let's look at a smaller example We can simulate the tossing of 4 fair coins
|
47
|
+
|
48
|
+
y = y = Bernoulli.new(4, 0.5)
|
49
|
+
# => #<Bernoulli:0x007ff894293198 @n=4, @p=0.5>
|
50
|
+
|
51
|
+
The method `sample` gives as a random array of length `n`, where each entry is `1` with a probability of `p`. The methods `sample_value` gives us the number of wins in a random expriment, we could than compare it to `expected_value`.
|
52
|
+
|
53
|
+
y.sample # => [1, 0, 1, 0]
|
54
|
+
y.sample # => [0, 0, 1, 0]
|
55
|
+
y.sample # => [1, 1, 1, 1]
|
56
|
+
|
57
|
+
y.sample_value # => 2
|
58
|
+
y.sample_value # => 1
|
59
|
+
y.sample_value # => 0
|
60
|
+
y.sample_value # => 2
|
61
|
+
|
62
|
+
y.expected_value # => 2.0
|
63
|
+
|
64
|
+
Last, but not least, we can create a full table of all possible values and their probabilitys like so
|
65
|
+
|
66
|
+
puts y.table
|
67
|
+
|
68
|
+
will produce
|
69
|
+
|
70
|
+
0,0.0625
|
71
|
+
1,0.25
|
72
|
+
2,0.375
|
73
|
+
3,0.25
|
74
|
+
4,0.0625
|
75
|
+
|
76
|
+
## Contributing
|
77
|
+
|
78
|
+
`bernoulli` is a really small project. After writing the same code for some project and then losing it two or three times I decided to do it one time and well, so I can just call in the code from here next time.
|
79
|
+
|
80
|
+
Feel free to cantact me about anything I could/should add or to contribute in any way to this simple library.
|
data/Rakefile
CHANGED
data/bernoulli.gemspec
CHANGED
@@ -1,13 +1,12 @@
|
|
1
|
-
|
2
1
|
Gem::Specification.new do |spec|
|
3
2
|
|
4
3
|
spec.name = 'bernoulli'
|
5
|
-
spec.version = '0.
|
4
|
+
spec.version = '0.2.0'
|
6
5
|
spec.summary = 'Binomial experiments library'
|
7
|
-
spec.description = 'A Library that allows calculation of probibilities and properties of binomial experiments'
|
8
|
-
spec.homepage = '
|
6
|
+
spec.description = 'A Library that allows calculation of probibilities and properties of binomial experiments like coin tossing'
|
7
|
+
spec.homepage = 'http://mkonutgan.github.com/bernoulli'
|
9
8
|
|
10
|
-
spec.date = '2012-05-
|
9
|
+
spec.date = '2012-05-28'
|
11
10
|
|
12
11
|
spec.author = 'Mikael Konutgan'
|
13
12
|
spec.email = 'mkonutgan@shortmail.com'
|
@@ -20,10 +19,12 @@ Gem::Specification.new do |spec|
|
|
20
19
|
'bernoulli.gemspec',
|
21
20
|
'lib/bernoulli.rb',
|
22
21
|
'lib/bernoulli/math.rb',
|
23
|
-
'test/test_bernoulli.rb'
|
22
|
+
'test/test_bernoulli.rb',
|
23
|
+
'test/test_math.rb'
|
24
24
|
]
|
25
25
|
spec.test_files = [
|
26
|
-
'test/test_bernoulli.rb'
|
26
|
+
'test/test_bernoulli.rb',
|
27
|
+
'test/test_math.rb'
|
27
28
|
]
|
28
29
|
|
29
30
|
end
|
data/lib/bernoulli/math.rb
CHANGED
data/lib/bernoulli.rb
CHANGED
@@ -1,5 +1,5 @@
|
|
1
|
-
|
2
1
|
require 'bernoulli/math'
|
2
|
+
require 'csv'
|
3
3
|
|
4
4
|
class Bernoulli
|
5
5
|
|
@@ -18,12 +18,12 @@ class Bernoulli
|
|
18
18
|
def [](k)
|
19
19
|
if k.is_a? Integer
|
20
20
|
raise Math::DomainError if not (k <= @n and k >= 0)
|
21
|
-
|
21
|
+
probability(k).to_f
|
22
22
|
elsif k.is_a? Range
|
23
23
|
raise Math::DomainError if not (k.begin >= 0 and k.end <= @n)
|
24
24
|
sum = 0
|
25
25
|
k.each do |i|
|
26
|
-
sum +=
|
26
|
+
sum += probability(i)
|
27
27
|
end
|
28
28
|
sum.to_f
|
29
29
|
else
|
@@ -31,6 +31,19 @@ class Bernoulli
|
|
31
31
|
end
|
32
32
|
end
|
33
33
|
|
34
|
+
def sample
|
35
|
+
s = []
|
36
|
+
@n.times do
|
37
|
+
s << (rand < @p ? 1 : 0)
|
38
|
+
end
|
39
|
+
s
|
40
|
+
end
|
41
|
+
|
42
|
+
def sample_value
|
43
|
+
sample.inject(:+)
|
44
|
+
end
|
45
|
+
alias :sv :sample_value
|
46
|
+
|
34
47
|
def expected_value
|
35
48
|
@n * @p
|
36
49
|
end
|
@@ -42,7 +55,7 @@ class Bernoulli
|
|
42
55
|
alias :v :variance
|
43
56
|
|
44
57
|
def standard_deviation
|
45
|
-
Math.sqrt(
|
58
|
+
Math.sqrt(variance)
|
46
59
|
end
|
47
60
|
alias :sd :standard_deviation
|
48
61
|
|
@@ -53,6 +66,14 @@ class Bernoulli
|
|
53
66
|
def excess
|
54
67
|
(1 - 6 * @p * (1 - @p)) / (@n * @p * (1 - @p))
|
55
68
|
end
|
69
|
+
|
70
|
+
def table
|
71
|
+
CSV.generate do |csv|
|
72
|
+
(0..@n).each do |n|
|
73
|
+
csv << [n, probability(n)]
|
74
|
+
end
|
75
|
+
end
|
76
|
+
end
|
56
77
|
|
57
78
|
end
|
58
79
|
|
data/test/test_bernoulli.rb
CHANGED
@@ -1,26 +1,8 @@
|
|
1
|
-
|
2
|
-
require '
|
1
|
+
require 'minitest/autorun'
|
2
|
+
require 'minitest/pride'
|
3
3
|
require 'bernoulli'
|
4
4
|
|
5
|
-
class BernoulliTest <
|
6
|
-
|
7
|
-
def test_factorial
|
8
|
-
assert_equal 1, Math.factorial(0)
|
9
|
-
assert_equal 1, Math.factorial(1)
|
10
|
-
assert_equal 120, Math.factorial(5)
|
11
|
-
assert_equal 263130836933693530167218012160000000, Math.factorial(32)
|
12
|
-
end
|
13
|
-
|
14
|
-
def test_binomial
|
15
|
-
assert_equal 1, Math.binomial(0, 0)
|
16
|
-
(1..10).each do |k|
|
17
|
-
assert_equal 1, Math.binomial(k, 0)
|
18
|
-
assert_equal 1, Math.binomial(k, k)
|
19
|
-
assert_equal k, Math.binomial(k, 1)
|
20
|
-
assert_equal k, Math.binomial(k, k - 1)
|
21
|
-
end
|
22
|
-
assert_equal 100947, Math.binomial(23, 17)
|
23
|
-
end
|
5
|
+
class BernoulliTest < MiniTest::Unit::TestCase
|
24
6
|
|
25
7
|
def test_new_experiment
|
26
8
|
assert_raises RuntimeError do
|
data/test/test_math.rb
ADDED
@@ -0,0 +1,39 @@
|
|
1
|
+
require 'minitest/autorun'
|
2
|
+
require 'minitest/pride'
|
3
|
+
require 'bernoulli'
|
4
|
+
|
5
|
+
class MathTest < MiniTest::Unit::TestCase
|
6
|
+
|
7
|
+
def test_factorial
|
8
|
+
assert_raises(Math::DomainError) { Math.factorial(-1) }
|
9
|
+
assert_raises(Math::DomainError) { Math.factorial(2.4) }
|
10
|
+
assert_equal 1, Math.factorial(0)
|
11
|
+
assert_equal 1, Math.factorial(1)
|
12
|
+
assert_equal 120, Math.factorial(5)
|
13
|
+
assert_equal 263130836933693530167218012160000000, Math.factorial(32)
|
14
|
+
end
|
15
|
+
|
16
|
+
def test_binomial
|
17
|
+
pairs = [[-1, 0], [2, 4], [-2, -1], [1.2, 3.4], [12, 50]]
|
18
|
+
pairs.each do |pair|
|
19
|
+
assert_raises Math::DomainError do
|
20
|
+
Math.binomial *pair
|
21
|
+
end
|
22
|
+
end
|
23
|
+
assert_equal 1, Math.binomial(0, 0)
|
24
|
+
(1..20).each do |k|
|
25
|
+
assert_equal 1, Math.binomial(k, 0)
|
26
|
+
assert_equal 1, Math.binomial(k, k)
|
27
|
+
assert_equal k, Math.binomial(k, 1)
|
28
|
+
assert_equal k, Math.binomial(k, k - 1)
|
29
|
+
end
|
30
|
+
(15..20).each do |n|
|
31
|
+
(5..15).each do |k|
|
32
|
+
assert_equal Math.binomial(n, k), Math.binomial(n, n - k)
|
33
|
+
end
|
34
|
+
end
|
35
|
+
assert_equal 100947, Math.binomial(23, 17)
|
36
|
+
end
|
37
|
+
|
38
|
+
end
|
39
|
+
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: bernoulli
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.
|
4
|
+
version: 0.2.0
|
5
5
|
prerelease:
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
@@ -9,10 +9,10 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2012-05-
|
12
|
+
date: 2012-05-28 00:00:00.000000000 Z
|
13
13
|
dependencies: []
|
14
14
|
description: A Library that allows calculation of probibilities and properties of
|
15
|
-
binomial experiments
|
15
|
+
binomial experiments like coin tossing
|
16
16
|
email: mkonutgan@shortmail.com
|
17
17
|
executables: []
|
18
18
|
extensions: []
|
@@ -24,7 +24,8 @@ files:
|
|
24
24
|
- lib/bernoulli.rb
|
25
25
|
- lib/bernoulli/math.rb
|
26
26
|
- test/test_bernoulli.rb
|
27
|
-
|
27
|
+
- test/test_math.rb
|
28
|
+
homepage: http://mkonutgan.github.com/bernoulli
|
28
29
|
licenses: []
|
29
30
|
post_install_message:
|
30
31
|
rdoc_options: []
|
@@ -50,3 +51,4 @@ specification_version: 3
|
|
50
51
|
summary: Binomial experiments library
|
51
52
|
test_files:
|
52
53
|
- test/test_bernoulli.rb
|
54
|
+
- test/test_math.rb
|