cmdstan 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +7 -0
- data/CHANGELOG.md +3 -0
- data/LICENSE.txt +29 -0
- data/README.md +84 -0
- data/ext/cmdstan/Makefile +5 -0
- data/ext/cmdstan/extconf.rb +60 -0
- data/lib/cmdstan.rb +20 -0
- data/lib/cmdstan/mcmc.rb +60 -0
- data/lib/cmdstan/mle.rb +21 -0
- data/lib/cmdstan/model.rb +107 -0
- data/lib/cmdstan/utils.rb +16 -0
- data/lib/cmdstan/version.rb +3 -0
- metadata +96 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA256:
|
3
|
+
metadata.gz: f52650c597cfa86c4e1bd0f095a2b321e906285df981f9f345bde5054fb8cc80
|
4
|
+
data.tar.gz: 3c98b6ff5c800ad9ecd28e325ed6f97d576306567a4ac8530bb117f948c6c321
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: ddab8b6b73175afcbee65f5653deee681b746fe2083015ec06b65866bc467ca2bb7ec6cc475a65bc83976ff4f13e8911ac08d999703c52e96d529f8b68408554
|
7
|
+
data.tar.gz: c8316343f62159e1963f7835bb231114275b15ab7fd982f3ae58f50c7168aad08e6448c6c2fd6110910f061595f586f3882a7a38a9e377c332b8b5a381bc4acb
|
data/CHANGELOG.md
ADDED
data/LICENSE.txt
ADDED
@@ -0,0 +1,29 @@
|
|
1
|
+
BSD 3-Clause License
|
2
|
+
|
3
|
+
Copyright (c) 2020, Andrew Kane
|
4
|
+
All rights reserved.
|
5
|
+
|
6
|
+
Redistribution and use in source and binary forms, with or without
|
7
|
+
modification, are permitted provided that the following conditions are met:
|
8
|
+
|
9
|
+
1. Redistributions of source code must retain the above copyright notice, this
|
10
|
+
list of conditions and the following disclaimer.
|
11
|
+
|
12
|
+
2. Redistributions in binary form must reproduce the above copyright notice,
|
13
|
+
this list of conditions and the following disclaimer in the documentation
|
14
|
+
and/or other materials provided with the distribution.
|
15
|
+
|
16
|
+
3. Neither the name of the copyright holder nor the names of its
|
17
|
+
contributors may be used to endorse or promote products derived from
|
18
|
+
this software without specific prior written permission.
|
19
|
+
|
20
|
+
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
|
21
|
+
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
22
|
+
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
|
23
|
+
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
|
24
|
+
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
25
|
+
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
|
26
|
+
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
|
27
|
+
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
|
28
|
+
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
29
|
+
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
data/README.md
ADDED
@@ -0,0 +1,84 @@
|
|
1
|
+
# CmdStan.rb
|
2
|
+
|
3
|
+
Bayesian inference for Ruby, powered by [CmdStan](https://github.com/stan-dev/cmdstan)
|
4
|
+
|
5
|
+
## Installation
|
6
|
+
|
7
|
+
Add this line to your application’s Gemfile:
|
8
|
+
|
9
|
+
```ruby
|
10
|
+
gem 'cmdstan'
|
11
|
+
```
|
12
|
+
|
13
|
+
Installation can take a few minutes as CmdStan downloads and builds.
|
14
|
+
|
15
|
+
## Getting Started
|
16
|
+
|
17
|
+
Create a Stan file, like `bernoulli.stan`
|
18
|
+
|
19
|
+
```stan
|
20
|
+
data {
|
21
|
+
int<lower=0> N;
|
22
|
+
int<lower=0,upper=1> y[N];
|
23
|
+
}
|
24
|
+
parameters {
|
25
|
+
real<lower=0,upper=1> theta;
|
26
|
+
}
|
27
|
+
model {
|
28
|
+
theta ~ beta(1,1);
|
29
|
+
y ~ bernoulli(theta);
|
30
|
+
}
|
31
|
+
```
|
32
|
+
|
33
|
+
Compile the model
|
34
|
+
|
35
|
+
```ruby
|
36
|
+
model = CmdStan::Model.new(stan_file: "bernoulli.stan")
|
37
|
+
```
|
38
|
+
|
39
|
+
Fit the model
|
40
|
+
|
41
|
+
```ruby
|
42
|
+
data = {"N" => 10, "y" => [0, 1, 0, 0, 0, 0, 0, 0, 0, 1]}
|
43
|
+
fit = model.sample(data: data, chains: 5)
|
44
|
+
```
|
45
|
+
|
46
|
+
Summarize the results
|
47
|
+
|
48
|
+
```ruby
|
49
|
+
fit.summary
|
50
|
+
```
|
51
|
+
|
52
|
+
## Maximum Likelihood Estimation
|
53
|
+
|
54
|
+
```ruby
|
55
|
+
mle = model.optimize(data: data)
|
56
|
+
mle.optimized_params
|
57
|
+
```
|
58
|
+
|
59
|
+
## Credits
|
60
|
+
|
61
|
+
This library is modeled after the [CmdStanPy API](https://github.com/stan-dev/cmdstanpy).
|
62
|
+
|
63
|
+
## History
|
64
|
+
|
65
|
+
View the [changelog](https://github.com/ankane/cmdstan/blob/master/CHANGELOG.md)
|
66
|
+
|
67
|
+
## Contributing
|
68
|
+
|
69
|
+
Everyone is encouraged to help improve this project. Here are a few ways you can help:
|
70
|
+
|
71
|
+
- [Report bugs](https://github.com/ankane/cmdstan/issues)
|
72
|
+
- Fix bugs and [submit pull requests](https://github.com/ankane/cmdstan/pulls)
|
73
|
+
- Write, clarify, or fix documentation
|
74
|
+
- Suggest or add new features
|
75
|
+
|
76
|
+
To get started with development:
|
77
|
+
|
78
|
+
```sh
|
79
|
+
git clone https://github.com/ankane/cmdstan.git
|
80
|
+
cd cmdstan
|
81
|
+
bundle install
|
82
|
+
ruby ext/cmdstan/extconf.rb
|
83
|
+
bundle exec rake test
|
84
|
+
```
|
@@ -0,0 +1,60 @@
|
|
1
|
+
require "digest"
|
2
|
+
require "fileutils"
|
3
|
+
require "net/http"
|
4
|
+
require "tmpdir"
|
5
|
+
|
6
|
+
version = "2.22.1"
|
7
|
+
checksum = "d12e46bda4bd3db9e8abe0554712b56e41f8e7843900338446d9a3b1acc2d0ce"
|
8
|
+
url = "https://github.com/stan-dev/cmdstan/releases/download/v#{version}/cmdstan-#{version}.tar.gz"
|
9
|
+
|
10
|
+
def download_file(url, download_path, checksum)
|
11
|
+
uri = URI(url)
|
12
|
+
location = nil
|
13
|
+
|
14
|
+
Net::HTTP.start(uri.host, uri.port, use_ssl: true) do |http|
|
15
|
+
request = Net::HTTP::Get.new(uri)
|
16
|
+
http.request(request) do |response|
|
17
|
+
case response
|
18
|
+
when Net::HTTPRedirection
|
19
|
+
location = response["location"]
|
20
|
+
when Net::HTTPSuccess
|
21
|
+
digest = Digest::SHA2.new
|
22
|
+
|
23
|
+
i = 0
|
24
|
+
File.open(download_path, "wb") do |f|
|
25
|
+
response.read_body do |chunk|
|
26
|
+
f.write(chunk)
|
27
|
+
digest.update(chunk)
|
28
|
+
|
29
|
+
# print progress
|
30
|
+
putc "." if i % 50 == 0
|
31
|
+
i += 1
|
32
|
+
end
|
33
|
+
end
|
34
|
+
puts # newline
|
35
|
+
|
36
|
+
abort "Bad checksum" if digest.hexdigest != checksum
|
37
|
+
else
|
38
|
+
abort "Bad response"
|
39
|
+
end
|
40
|
+
end
|
41
|
+
end
|
42
|
+
|
43
|
+
# outside of Net::HTTP block to close previous connection
|
44
|
+
download_file(location, download_path, checksum) if location
|
45
|
+
end
|
46
|
+
|
47
|
+
# download
|
48
|
+
puts "Downloading #{url}..."
|
49
|
+
download_path = "#{Dir.tmpdir}/cmdstan-#{version}.tar.gz"
|
50
|
+
download_file(url, download_path, checksum)
|
51
|
+
|
52
|
+
# extract
|
53
|
+
path = ENV["CMDSTAN"] || File.expand_path("../../tmp/cmdstan", __dir__)
|
54
|
+
FileUtils.mkdir_p(path)
|
55
|
+
Dir.chdir(path)
|
56
|
+
# TODO use Gem::Package::TarReader from Rubygems
|
57
|
+
system "tar", "zxvf", download_path, "-C", path, "--strip-components=1"
|
58
|
+
|
59
|
+
# build
|
60
|
+
system "make", "build", "-j"
|
data/lib/cmdstan.rb
ADDED
@@ -0,0 +1,20 @@
|
|
1
|
+
# stdlib
|
2
|
+
require "csv"
|
3
|
+
require "json"
|
4
|
+
require "open3"
|
5
|
+
|
6
|
+
# modules
|
7
|
+
require "cmdstan/utils"
|
8
|
+
require "cmdstan/mcmc"
|
9
|
+
require "cmdstan/mle"
|
10
|
+
require "cmdstan/model"
|
11
|
+
require "cmdstan/version"
|
12
|
+
|
13
|
+
module CmdStan
|
14
|
+
class Error < StandardError; end
|
15
|
+
|
16
|
+
class << self
|
17
|
+
attr_accessor :path
|
18
|
+
end
|
19
|
+
self.path = ENV["CMDSTAN"] || File.expand_path("../tmp/cmdstan", __dir__)
|
20
|
+
end
|
data/lib/cmdstan/mcmc.rb
ADDED
@@ -0,0 +1,60 @@
|
|
1
|
+
module CmdStan
|
2
|
+
class MCMC
|
3
|
+
include Utils
|
4
|
+
|
5
|
+
attr_reader :column_names, :draws
|
6
|
+
|
7
|
+
# private
|
8
|
+
# TODO use runset for args
|
9
|
+
def initialize(output_files)
|
10
|
+
@output_files = output_files
|
11
|
+
validate_csv_files
|
12
|
+
end
|
13
|
+
|
14
|
+
def sample
|
15
|
+
sample = []
|
16
|
+
@output_files.each_with_index do |output_file, chain|
|
17
|
+
i = 0
|
18
|
+
CSV.foreach(output_file.path, skip_lines: /^#/, headers: true) do |row|
|
19
|
+
(sample[i] ||= [])[chain] = row.to_h.values.map(&:to_f)
|
20
|
+
i += 1
|
21
|
+
end
|
22
|
+
raise "Bug detected" if i != draws
|
23
|
+
end
|
24
|
+
sample
|
25
|
+
end
|
26
|
+
|
27
|
+
def summary
|
28
|
+
csv_file = Tempfile.new
|
29
|
+
run_command "#{CmdStan.path}/bin/stansummary", "--csv_file=#{csv_file.path}", *@output_files.map(&:path)
|
30
|
+
|
31
|
+
result = {}
|
32
|
+
CSV.foreach(csv_file.path, headers: true, converters: :numeric) do |row|
|
33
|
+
value = row.to_h
|
34
|
+
name = value.delete("name")
|
35
|
+
result[name] = value if name == "lp__" || !name.end_with?("__")
|
36
|
+
end
|
37
|
+
result
|
38
|
+
end
|
39
|
+
|
40
|
+
private
|
41
|
+
|
42
|
+
def validate_csv_files
|
43
|
+
# TODO ensure consistent files
|
44
|
+
output_file = @output_files.first
|
45
|
+
|
46
|
+
File.foreach(output_file.path) do |line|
|
47
|
+
matches = /num_samples = (\d+)/.match(line)
|
48
|
+
if matches
|
49
|
+
@draws = matches[1].to_i
|
50
|
+
break
|
51
|
+
end
|
52
|
+
end
|
53
|
+
|
54
|
+
CSV.foreach(output_file.path, skip_lines: /^#/, headers: true) do |row|
|
55
|
+
@column_names = row.to_h.keys
|
56
|
+
break
|
57
|
+
end
|
58
|
+
end
|
59
|
+
end
|
60
|
+
end
|
data/lib/cmdstan/mle.rb
ADDED
@@ -0,0 +1,21 @@
|
|
1
|
+
module CmdStan
|
2
|
+
class MLE
|
3
|
+
include Utils
|
4
|
+
|
5
|
+
# private
|
6
|
+
# TODO use runset for args
|
7
|
+
def initialize(output_file)
|
8
|
+
@output_file = output_file
|
9
|
+
end
|
10
|
+
|
11
|
+
def optimized_params
|
12
|
+
CSV.foreach(@output_file.path, skip_lines: /^#/, headers: true, converters: :numeric) do |row|
|
13
|
+
return row.to_h
|
14
|
+
end
|
15
|
+
end
|
16
|
+
|
17
|
+
def column_names
|
18
|
+
optimized_params.keys
|
19
|
+
end
|
20
|
+
end
|
21
|
+
end
|
@@ -0,0 +1,107 @@
|
|
1
|
+
module CmdStan
|
2
|
+
class Model
|
3
|
+
include Utils
|
4
|
+
|
5
|
+
attr_reader :exe_file, :name, :stan_file
|
6
|
+
|
7
|
+
def initialize(stan_file: nil, exe_file: nil, compile: true)
|
8
|
+
# convert to absolute path
|
9
|
+
stan_file = File.expand_path(stan_file) if stan_file
|
10
|
+
|
11
|
+
@stan_file = stan_file
|
12
|
+
@exe_file = exe_file || stan_file.sub(/.stan\z/, "")
|
13
|
+
@name = File.basename(@exe_file)
|
14
|
+
|
15
|
+
if compile && !exe_file
|
16
|
+
self.compile
|
17
|
+
end
|
18
|
+
end
|
19
|
+
|
20
|
+
def compile
|
21
|
+
Dir.chdir(CmdStan.path) do
|
22
|
+
run_command "make", @exe_file
|
23
|
+
end
|
24
|
+
end
|
25
|
+
|
26
|
+
def code
|
27
|
+
File.read(stan_file)
|
28
|
+
end
|
29
|
+
|
30
|
+
def sample(data:, chains: nil, seed: nil, inits: nil, warmup_iters: nil, sampling_iters: nil)
|
31
|
+
data_file = Tempfile.new(["cmdstan", ".json"])
|
32
|
+
data_file.write(data.to_json)
|
33
|
+
data_file.close
|
34
|
+
|
35
|
+
chain ||= 4
|
36
|
+
|
37
|
+
output_files = []
|
38
|
+
chains.times do |chain|
|
39
|
+
output_file = Tempfile.new(["cmdstan", ".csv"])
|
40
|
+
|
41
|
+
args = [@exe_file, "id=#{chain + 1}"]
|
42
|
+
|
43
|
+
# random
|
44
|
+
args += ["random", "seed=#{seed.to_i}"] if seed
|
45
|
+
|
46
|
+
# data
|
47
|
+
args += ["data", "file=#{data_file.path}"]
|
48
|
+
if inits
|
49
|
+
init_file = Tempfile.new(["cmdstan", ".json"])
|
50
|
+
init_file.write(inits.to_json)
|
51
|
+
init_file.close
|
52
|
+
args << "init=#{init_file.path}"
|
53
|
+
end
|
54
|
+
|
55
|
+
# output
|
56
|
+
args += ["output", "file=#{output_file.path}"]
|
57
|
+
|
58
|
+
# method
|
59
|
+
args += ["method=sample"]
|
60
|
+
args << "num_warmup=#{warmup_iters.to_i}" if warmup_iters
|
61
|
+
args << "num_samples=#{sampling_iters.to_i}" if sampling_iters
|
62
|
+
args += ["algorithm=hmc", "adapt", "engaged=1"]
|
63
|
+
|
64
|
+
run_command *args
|
65
|
+
|
66
|
+
output_files << output_file
|
67
|
+
end
|
68
|
+
|
69
|
+
MCMC.new(output_files)
|
70
|
+
end
|
71
|
+
|
72
|
+
def optimize(data:, seed: nil, inits: nil, algorithm: nil, iter: nil)
|
73
|
+
data_file = Tempfile.new(["cmdstan", ".json"])
|
74
|
+
data_file.write(data.to_json)
|
75
|
+
data_file.close
|
76
|
+
|
77
|
+
output_file = Tempfile.new(["cmdstan", ".csv"])
|
78
|
+
diagnostic_file = Tempfile.new(["cmdstan", ".csv"])
|
79
|
+
|
80
|
+
args = [@exe_file]
|
81
|
+
|
82
|
+
# random
|
83
|
+
args += ["random", "seed=#{seed.to_i}"] if seed
|
84
|
+
|
85
|
+
# data
|
86
|
+
args += ["data", "file=#{data_file.path}"]
|
87
|
+
if inits
|
88
|
+
init_file = Tempfile.new(["cmdstan", ".json"])
|
89
|
+
init_file.write(inits.to_json)
|
90
|
+
init_file.close
|
91
|
+
args << "init=#{init_file.path}"
|
92
|
+
end
|
93
|
+
|
94
|
+
# output
|
95
|
+
args += ["output", "file=#{output_file.path}", "diagnostic_file=#{diagnostic_file.path}"]
|
96
|
+
|
97
|
+
# method
|
98
|
+
args << "method=optimize"
|
99
|
+
args << "algorithm=#{algorithm.to_s.downcase}" if algorithm
|
100
|
+
args << "iter=#{iter.to_i}" if iter
|
101
|
+
|
102
|
+
run_command *args
|
103
|
+
|
104
|
+
MLE.new(output_file)
|
105
|
+
end
|
106
|
+
end
|
107
|
+
end
|
@@ -0,0 +1,16 @@
|
|
1
|
+
module CmdStan
|
2
|
+
module Utils
|
3
|
+
private
|
4
|
+
|
5
|
+
def run_command(*args)
|
6
|
+
# use popen3 since it does escaping (like system)
|
7
|
+
Open3.popen3(*args) do |i, o, e, t|
|
8
|
+
if t.value.exitstatus != 0
|
9
|
+
$stderr.puts o.read
|
10
|
+
$stderr.puts e.read
|
11
|
+
raise Error, "Command failed"
|
12
|
+
end
|
13
|
+
end
|
14
|
+
end
|
15
|
+
end
|
16
|
+
end
|
metadata
ADDED
@@ -0,0 +1,96 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: cmdstan
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.1.0
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Andrew Kane
|
8
|
+
autorequire:
|
9
|
+
bindir: bin
|
10
|
+
cert_chain: []
|
11
|
+
date: 2020-04-09 00:00:00.000000000 Z
|
12
|
+
dependencies:
|
13
|
+
- !ruby/object:Gem::Dependency
|
14
|
+
name: bundler
|
15
|
+
requirement: !ruby/object:Gem::Requirement
|
16
|
+
requirements:
|
17
|
+
- - ">="
|
18
|
+
- !ruby/object:Gem::Version
|
19
|
+
version: '0'
|
20
|
+
type: :development
|
21
|
+
prerelease: false
|
22
|
+
version_requirements: !ruby/object:Gem::Requirement
|
23
|
+
requirements:
|
24
|
+
- - ">="
|
25
|
+
- !ruby/object:Gem::Version
|
26
|
+
version: '0'
|
27
|
+
- !ruby/object:Gem::Dependency
|
28
|
+
name: rake
|
29
|
+
requirement: !ruby/object:Gem::Requirement
|
30
|
+
requirements:
|
31
|
+
- - ">="
|
32
|
+
- !ruby/object:Gem::Version
|
33
|
+
version: '0'
|
34
|
+
type: :development
|
35
|
+
prerelease: false
|
36
|
+
version_requirements: !ruby/object:Gem::Requirement
|
37
|
+
requirements:
|
38
|
+
- - ">="
|
39
|
+
- !ruby/object:Gem::Version
|
40
|
+
version: '0'
|
41
|
+
- !ruby/object:Gem::Dependency
|
42
|
+
name: minitest
|
43
|
+
requirement: !ruby/object:Gem::Requirement
|
44
|
+
requirements:
|
45
|
+
- - ">="
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: '5'
|
48
|
+
type: :development
|
49
|
+
prerelease: false
|
50
|
+
version_requirements: !ruby/object:Gem::Requirement
|
51
|
+
requirements:
|
52
|
+
- - ">="
|
53
|
+
- !ruby/object:Gem::Version
|
54
|
+
version: '5'
|
55
|
+
description:
|
56
|
+
email: andrew@chartkick.com
|
57
|
+
executables: []
|
58
|
+
extensions:
|
59
|
+
- ext/cmdstan/extconf.rb
|
60
|
+
extra_rdoc_files: []
|
61
|
+
files:
|
62
|
+
- CHANGELOG.md
|
63
|
+
- LICENSE.txt
|
64
|
+
- README.md
|
65
|
+
- ext/cmdstan/Makefile
|
66
|
+
- ext/cmdstan/extconf.rb
|
67
|
+
- lib/cmdstan.rb
|
68
|
+
- lib/cmdstan/mcmc.rb
|
69
|
+
- lib/cmdstan/mle.rb
|
70
|
+
- lib/cmdstan/model.rb
|
71
|
+
- lib/cmdstan/utils.rb
|
72
|
+
- lib/cmdstan/version.rb
|
73
|
+
homepage: https://github.com/ankane/cmdstan
|
74
|
+
licenses:
|
75
|
+
- BSD-3-Clause
|
76
|
+
metadata: {}
|
77
|
+
post_install_message:
|
78
|
+
rdoc_options: []
|
79
|
+
require_paths:
|
80
|
+
- lib
|
81
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
82
|
+
requirements:
|
83
|
+
- - ">="
|
84
|
+
- !ruby/object:Gem::Version
|
85
|
+
version: '2.4'
|
86
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
87
|
+
requirements:
|
88
|
+
- - ">="
|
89
|
+
- !ruby/object:Gem::Version
|
90
|
+
version: '0'
|
91
|
+
requirements: []
|
92
|
+
rubygems_version: 3.1.2
|
93
|
+
signing_key:
|
94
|
+
specification_version: 4
|
95
|
+
summary: Bayesian inference for Ruby, powered by CmdStan
|
96
|
+
test_files: []
|