Merge remote-tracking branch 'upstream/master'

This commit is contained in:
Ryan Boldi 2023-03-17 10:45:37 -04:00
commit 1c0805169f
139 changed files with 1003581 additions and 314 deletions

View File

@ -0,0 +1,24 @@
{:input1 1, :output1 "1"}
{:input1 2, :output1 "2"}
{:input1 3, :output1 "Fizz"}
{:input1 4, :output1 "4"}
{:input1 5, :output1 "Buzz"}
{:input1 6, :output1 "Fizz"}
{:input1 7, :output1 "7"}
{:input1 8, :output1 "8"}
{:input1 9, :output1 "Fizz"}
{:input1 10, :output1 "Buzz"}
{:input1 11, :output1 "11"}
{:input1 12, :output1 "Fizz"}
{:input1 13, :output1 "13"}
{:input1 14, :output1 "14"}
{:input1 15, :output1 "FizzBuzz"}
{:input1 16, :output1 "16"}
{:input1 17, :output1 "17"}
{:input1 18, :output1 "Fizz"}
{:input1 19, :output1 "19"}
{:input1 20, :output1 "Buzz"}
{:input1 49995, :output1 "FizzBuzz"}
{:input1 49998, :output1 "Fizz"}
{:input1 49999, :output1 "49999"}
{:input1 50000, :output1 "Buzz"}

File diff suppressed because it is too large Load Diff

View File

@ -2,12 +2,27 @@
Yet another Push-based genetic programming system in Clojure.
Full documentation is on the GitHub pages link.
## Usage
If you have installed [leiningen](https://leiningen.org), which is a tool
for running Clojure programs, then you can run Propeller on a genetic
programming problem that is defined within this project from the command
line with the command `lein run -m <namespace>`, replacing `<namespace>`
If you are working in a Clojure IDE with an integrated REPL, the first
thing you may want to do is to open `src/propeller/session.cljc` and
evaluate the namespace declaration and the commented-out expressions
therein. These demonstrate core components of Propeller including
complete genetic programming runs.
To run Propeller from the command line, on a genetic programming problem
that is defined within this project, you will probably want to use either
the Clojure [CLI tools](https://clojure.org/guides/deps_and_cli) or
[leiningen](https://leiningen.org).
The instructions below are written for leiningen. If you are using
the CLI tools instead, then replace `lein run -m` in each command
with `clj -M -m`.
If you are using leiningen, then you can start a run with the command
`lein run -m <namespace>`, replacing `<namespace>`
with the actual namespace that you will find at the top of the problem file.
For example, you can run the simple-regression genetic programming problem with:
@ -44,22 +59,6 @@ containing curly brackets that may confuse your shell:
lein run -m propeller.problems.simple-regression :variation "{:umad 1.0}"
```
To run a genetic programming problem from a REPL, start
your REPL for the project (e.g. with `lein repl` at the
command line when in the project directory, or through your
IDE) and then do something like the following (which in
this case runs the simple-regression problem with
`:population-size` 100):
```
(require 'propeller.problems.simple-regression)
(in-ns 'propeller.problems.simple-regression)
(-main :population-size 100 :variation {:umad 1.0})
```
If you want to run the problem with the default parameters,
then you should call `-main` without arguments, as `(-main)`.
## CLJS Usage

View File

@ -4,4 +4,10 @@
org.clojure/clojurescript #:mvn{:version "1.9.946"},
org.clojure/test.check #:mvn{:version "1.1.0"},
net.clojars.schneau/psb2 #:mvn{:version "1.1.0"}},
:mvn/repos {}}
:mvn/repos {}
:codox {:extra-deps {codox/codox {:mvn/version "0.10.8"}}
:exec-fn codox.main/generate-docs
:exec-args {:doc-paths ["src/docs_src"]
:output-path "docs"
:metadata {:doc "FIXME: write docs" :doc/format :markdown}}
}}

View File

@ -1,32 +0,0 @@
# Introduction to Propeller
TODO: write [great documentation](http://jacobian.org/writing/what-to-write/)
# Simplification
To use Propeller's auto-simplification system, simply include the following four command line arguments when running a problem:
```clojure
:simplification? true
```
Toggle auto-simplification
```clojure
:simplification-k 4
```
This is the upper bound for elements deleted from the plushy every step. Every step, a number in $[1, k]$ of elements is deleted from the plushy representation of the solution.
```clojure
:simplification-steps 1000
```
Number of simplification steps to perform
```clojure
:simplification-verbose? true
```
whether or not to output simplification info into the output of the evolutionary run.
The output with verbose adds the following lines to the output:
```clojure
{:start-plushy-length 42, :k 4}
{:final-plushy-length 13, :final-plushy (:in1 :in1 :integer_quot :in1 :in1 :exec_dup :in1 :integer_mult close :exec_dup :integer_add 1 :integer_add)}
```

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

117
docs/Adding_Problem.html Normal file

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

14
docs/Genome.html Normal file

File diff suppressed because one or more lines are too long

134
docs/Intro.html Normal file

File diff suppressed because one or more lines are too long

15
docs/Selection.html Normal file

File diff suppressed because one or more lines are too long

20
docs/Simplification.html Normal file

File diff suppressed because one or more lines are too long

38
docs/Variation.html Normal file

File diff suppressed because one or more lines are too long

551
docs/css/default.css Normal file
View File

@ -0,0 +1,551 @@
body {
font-family: Helvetica, Arial, sans-serif;
font-size: 15px;
}
pre, code {
font-family: Monaco, DejaVu Sans Mono, Consolas, monospace;
font-size: 9pt;
margin: 15px 0;
}
h1 {
font-weight: normal;
font-size: 29px;
margin: 10px 0 2px 0;
padding: 0;
}
h2 {
font-weight: normal;
font-size: 25px;
}
h5.license {
margin: 9px 0 22px 0;
color: #555;
font-weight: normal;
font-size: 12px;
font-style: italic;
}
.document h1, .namespace-index h1 {
font-size: 32px;
margin-top: 12px;
}
#header, #content, .sidebar {
position: fixed;
}
#header {
top: 0;
left: 0;
right: 0;
height: 22px;
color: #f5f5f5;
padding: 5px 7px;
}
#content {
top: 32px;
right: 0;
bottom: 0;
overflow: auto;
background: #fff;
color: #333;
padding: 0 18px;
}
.sidebar {
position: fixed;
top: 32px;
bottom: 0;
overflow: auto;
}
.sidebar.primary {
background: #e2e2e2;
border-right: solid 1px #cccccc;
left: 0;
width: 250px;
}
.sidebar.secondary {
background: #f2f2f2;
border-right: solid 1px #d7d7d7;
left: 251px;
width: 200px;
}
#content.namespace-index, #content.document {
left: 251px;
}
#content.namespace-docs {
left: 452px;
}
#content.document {
padding-bottom: 10%;
}
#header {
background: #3f3f3f;
box-shadow: 0 0 8px rgba(0, 0, 0, 0.4);
z-index: 100;
}
#header h1 {
margin: 0;
padding: 0;
font-size: 18px;
font-weight: lighter;
text-shadow: -1px -1px 0px #333;
}
#header h1 .project-version {
font-weight: normal;
}
.project-version {
padding-left: 0.15em;
}
#header a, .sidebar a {
display: block;
text-decoration: none;
}
#header a {
color: #f5f5f5;
}
.sidebar a {
color: #333;
}
#header h2 {
float: right;
font-size: 9pt;
font-weight: normal;
margin: 4px 3px;
padding: 0;
color: #bbb;
}
#header h2 a {
display: inline;
}
.sidebar h3 {
margin: 0;
padding: 10px 13px 0 13px;
font-size: 19px;
font-weight: lighter;
}
.sidebar h3 a {
color: #444;
}
.sidebar h3.no-link {
color: #636363;
}
.sidebar ul {
padding: 7px 0 6px 0;
margin: 0;
}
.sidebar ul.index-link {
padding-bottom: 4px;
}
.sidebar li {
display: block;
vertical-align: middle;
}
.sidebar li a, .sidebar li .no-link {
border-left: 3px solid transparent;
padding: 0 10px;
white-space: nowrap;
}
.sidebar li .no-link {
display: block;
color: #777;
font-style: italic;
}
.sidebar li .inner {
display: inline-block;
padding-top: 7px;
height: 24px;
}
.sidebar li a, .sidebar li .tree {
height: 31px;
}
.depth-1 .inner { padding-left: 2px; }
.depth-2 .inner { padding-left: 6px; }
.depth-3 .inner { padding-left: 20px; }
.depth-4 .inner { padding-left: 34px; }
.depth-5 .inner { padding-left: 48px; }
.depth-6 .inner { padding-left: 62px; }
.sidebar li .tree {
display: block;
float: left;
position: relative;
top: -10px;
margin: 0 4px 0 0;
padding: 0;
}
.sidebar li.depth-1 .tree {
display: none;
}
.sidebar li .tree .top, .sidebar li .tree .bottom {
display: block;
margin: 0;
padding: 0;
width: 7px;
}
.sidebar li .tree .top {
border-left: 1px solid #aaa;
border-bottom: 1px solid #aaa;
height: 19px;
}
.sidebar li .tree .bottom {
height: 22px;
}
.sidebar li.branch .tree .bottom {
border-left: 1px solid #aaa;
}
.sidebar.primary li.current a {
border-left: 3px solid #a33;
color: #a33;
}
.sidebar.secondary li.current a {
border-left: 3px solid #33a;
color: #33a;
}
.namespace-index h2 {
margin: 30px 0 0 0;
}
.namespace-index h3 {
font-size: 16px;
font-weight: bold;
margin-bottom: 0;
}
.namespace-index .topics {
padding-left: 30px;
margin: 11px 0 0 0;
}
.namespace-index .topics li {
padding: 5px 0;
}
.namespace-docs h3 {
font-size: 18px;
font-weight: bold;
}
.public h3 {
margin: 0;
float: left;
}
.usage {
clear: both;
}
.public {
margin: 0;
border-top: 1px solid #e0e0e0;
padding-top: 14px;
padding-bottom: 6px;
}
.public:last-child {
margin-bottom: 20%;
}
.members .public:last-child {
margin-bottom: 0;
}
.members {
margin: 15px 0;
}
.members h4 {
color: #555;
font-weight: normal;
font-variant: small-caps;
margin: 0 0 5px 0;
}
.members .inner {
padding-top: 5px;
padding-left: 12px;
margin-top: 2px;
margin-left: 7px;
border-left: 1px solid #bbb;
}
#content .members .inner h3 {
font-size: 12pt;
}
.members .public {
border-top: none;
margin-top: 0;
padding-top: 6px;
padding-bottom: 0;
}
.members .public:first-child {
padding-top: 0;
}
h4.type,
h4.dynamic,
h4.added,
h4.deprecated {
float: left;
margin: 3px 10px 15px 0;
font-size: 15px;
font-weight: bold;
font-variant: small-caps;
}
.public h4.type,
.public h4.dynamic,
.public h4.added,
.public h4.deprecated {
font-size: 13px;
font-weight: bold;
margin: 3px 0 0 10px;
}
.members h4.type,
.members h4.added,
.members h4.deprecated {
margin-top: 1px;
}
h4.type {
color: #717171;
}
h4.dynamic {
color: #9933aa;
}
h4.added {
color: #508820;
}
h4.deprecated {
color: #880000;
}
.namespace {
margin-bottom: 30px;
}
.namespace:last-child {
margin-bottom: 10%;
}
.index {
padding: 0;
font-size: 80%;
margin: 15px 0;
line-height: 16px;
}
.index * {
display: inline;
}
.index p {
padding-right: 3px;
}
.index li {
padding-right: 5px;
}
.index ul {
padding-left: 0;
}
.type-sig {
clear: both;
color: #088;
}
.type-sig pre {
padding-top: 10px;
margin: 0;
}
.usage code {
display: block;
color: #008;
margin: 2px 0;
}
.usage code:first-child {
padding-top: 10px;
}
p {
margin: 15px 0;
}
.public p:first-child, .public pre.plaintext {
margin-top: 12px;
}
.doc {
margin: 0 0 26px 0;
clear: both;
}
.public .doc {
margin: 0;
}
.namespace-index .doc {
margin-bottom: 20px;
}
.namespace-index .namespace .doc {
margin-bottom: 10px;
}
.markdown p, .markdown li, .markdown dt, .markdown dd, .markdown td {
line-height: 22px;
}
.markdown li {
padding: 2px 0;
}
.markdown h2 {
font-weight: normal;
font-size: 25px;
margin: 30px 0 10px 0;
}
.markdown h3 {
font-weight: normal;
font-size: 20px;
margin: 30px 0 0 0;
}
.markdown h4 {
font-size: 15px;
margin: 22px 0 -4px 0;
}
.doc, .public, .namespace .index {
max-width: 680px;
overflow-x: visible;
}
.markdown pre > code {
display: block;
padding: 10px;
}
.markdown pre > code, .src-link a {
border: 1px solid #e4e4e4;
border-radius: 2px;
}
.markdown code:not(.hljs), .src-link a {
background: #f6f6f6;
}
pre.deps {
display: inline-block;
margin: 0 10px;
border: 1px solid #e4e4e4;
border-radius: 2px;
padding: 10px;
background-color: #f6f6f6;
}
.markdown hr {
border-style: solid;
border-top: none;
color: #ccc;
}
.doc ul, .doc ol {
padding-left: 30px;
}
.doc table {
border-collapse: collapse;
margin: 0 10px;
}
.doc table td, .doc table th {
border: 1px solid #dddddd;
padding: 4px 6px;
}
.doc table th {
background: #f2f2f2;
}
.doc dl {
margin: 0 10px 20px 10px;
}
.doc dl dt {
font-weight: bold;
margin: 0;
padding: 3px 0;
border-bottom: 1px solid #ddd;
}
.doc dl dd {
padding: 5px 0;
margin: 0 0 5px 10px;
}
.doc abbr {
border-bottom: 1px dotted #333;
font-variant: none;
cursor: help;
}
.src-link {
margin-bottom: 15px;
}
.src-link a {
font-size: 70%;
padding: 1px 4px;
text-decoration: none;
color: #5555bb;
}

97
docs/css/highlight.css Normal file
View File

@ -0,0 +1,97 @@
/*
github.com style (c) Vasily Polovnyov <vast@whiteants.net>
*/
.hljs {
display: block;
overflow-x: auto;
padding: 0.5em;
color: #333;
background: #f8f8f8;
}
.hljs-comment,
.hljs-quote {
color: #998;
font-style: italic;
}
.hljs-keyword,
.hljs-selector-tag,
.hljs-subst {
color: #333;
font-weight: bold;
}
.hljs-number,
.hljs-literal,
.hljs-variable,
.hljs-template-variable,
.hljs-tag .hljs-attr {
color: #008080;
}
.hljs-string,
.hljs-doctag {
color: #d14;
}
.hljs-title,
.hljs-section,
.hljs-selector-id {
color: #900;
font-weight: bold;
}
.hljs-subst {
font-weight: normal;
}
.hljs-type,
.hljs-class .hljs-title {
color: #458;
font-weight: bold;
}
.hljs-tag,
.hljs-name,
.hljs-attribute {
color: #000080;
font-weight: normal;
}
.hljs-regexp,
.hljs-link {
color: #009926;
}
.hljs-symbol,
.hljs-bullet {
color: #990073;
}
.hljs-built_in,
.hljs-builtin-name {
color: #0086b3;
}
.hljs-meta {
color: #999;
font-weight: bold;
}
.hljs-deletion {
background: #fdd;
}
.hljs-addition {
background: #dfd;
}
.hljs-emphasis {
font-style: italic;
}
.hljs-strong {
font-weight: bold;
}

49
docs/index.html Normal file

File diff suppressed because one or more lines are too long

2
docs/js/highlight.min.js vendored Normal file

File diff suppressed because one or more lines are too long

4
docs/js/jquery.min.js vendored Normal file

File diff suppressed because one or more lines are too long

112
docs/js/page_effects.js Normal file
View File

@ -0,0 +1,112 @@
function visibleInParent(element) {
var position = $(element).position().top
return position > -50 && position < ($(element).offsetParent().height() - 50)
}
function hasFragment(link, fragment) {
return $(link).attr("href").indexOf("#" + fragment) != -1
}
function findLinkByFragment(elements, fragment) {
return $(elements).filter(function(i, e) { return hasFragment(e, fragment)}).first()
}
function scrollToCurrentVarLink(elements) {
var elements = $(elements);
var parent = elements.offsetParent();
if (elements.length == 0) return;
var top = elements.first().position().top;
var bottom = elements.last().position().top + elements.last().height();
if (top >= 0 && bottom <= parent.height()) return;
if (top < 0) {
parent.scrollTop(parent.scrollTop() + top);
}
else if (bottom > parent.height()) {
parent.scrollTop(parent.scrollTop() + bottom - parent.height());
}
}
function setCurrentVarLink() {
$('.secondary a').parent().removeClass('current')
$('.anchor').
filter(function(index) { return visibleInParent(this) }).
each(function(index, element) {
findLinkByFragment(".secondary a", element.id).
parent().
addClass('current')
});
scrollToCurrentVarLink('.secondary .current');
}
var hasStorage = (function() { try { return localStorage.getItem } catch(e) {} }())
function scrollPositionId(element) {
var directory = window.location.href.replace(/[^\/]+\.html$/, '')
return 'scroll::' + $(element).attr('id') + '::' + directory
}
function storeScrollPosition(element) {
if (!hasStorage) return;
localStorage.setItem(scrollPositionId(element) + "::x", $(element).scrollLeft())
localStorage.setItem(scrollPositionId(element) + "::y", $(element).scrollTop())
}
function recallScrollPosition(element) {
if (!hasStorage) return;
$(element).scrollLeft(localStorage.getItem(scrollPositionId(element) + "::x"))
$(element).scrollTop(localStorage.getItem(scrollPositionId(element) + "::y"))
}
function persistScrollPosition(element) {
recallScrollPosition(element)
$(element).scroll(function() { storeScrollPosition(element) })
}
function sidebarContentWidth(element) {
var widths = $(element).find('.inner').map(function() { return $(this).innerWidth() })
return Math.max.apply(Math, widths)
}
function calculateSize(width, snap, margin, minimum) {
if (width == 0) {
return 0
}
else {
return Math.max(minimum, (Math.ceil(width / snap) * snap) + (margin * 2))
}
}
function resizeSidebars() {
var primaryWidth = sidebarContentWidth('.primary')
var secondaryWidth = 0
if ($('.secondary').length != 0) {
secondaryWidth = sidebarContentWidth('.secondary')
}
// snap to grid
primaryWidth = calculateSize(primaryWidth, 32, 13, 160)
secondaryWidth = calculateSize(secondaryWidth, 32, 13, 160)
$('.primary').css('width', primaryWidth)
$('.secondary').css('width', secondaryWidth).css('left', primaryWidth + 1)
if (secondaryWidth > 0) {
$('#content').css('left', primaryWidth + secondaryWidth + 2)
}
else {
$('#content').css('left', primaryWidth + 1)
}
}
$(window).ready(resizeSidebars)
$(window).ready(setCurrentVarLink)
$(window).ready(function() { persistScrollPosition('.primary')})
$(window).ready(function() {
$('#content').scroll(setCurrentVarLink)
$(window).resize(setCurrentVarLink)
})

4
docs/propeller.core.html Normal file

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

9
docs/propeller.gp.html Normal file

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

12
docs/propeller.utils.html Normal file

File diff suppressed because one or more lines are too long

File diff suppressed because one or more lines are too long

View File

@ -11,4 +11,8 @@
:profiles {:profiling {:dependencies [[com.clojure-goes-fast/clj-async-profiler "0.5.1"]]}}
:main ^:skip-aot propeller.core
:repl-options {:init-ns propeller.core}
:jvm-opts ^:replace [])
:jvm-opts ^:replace []
:plugins [[lein-codox "0.10.8"]]
:codox {:output-path "docs"
:metadata {:doc "FIXME: write docs" :doc/format :markdown}
:doc-paths ["src/docs_src"]})

73
scripts/FunctionsToMD.py Normal file
View File

@ -0,0 +1,73 @@
# Takes Push instructions defined through (def-instruction and
# puts their documentation into a markdown file in docs_src
import os
from mdutils.mdutils import MdUtils
mdFile = MdUtils(file_name='src/docs_src/Additional_Instructions')
mdFile.new_header(level=1, title='Additional Instructions')
os.chdir('..')
instructionFiles = os.listdir('src/propeller/push/instructions')
instructionFiles.remove('vector.cljc')
instructionFiles.remove('polymorphic.cljc')
print(instructionFiles)
hasDefInstruction = False
for file in instructionFiles:
mdFile.new_header(level=1, title=file)
try:
print(file)
# opening and reading the file
file_read = open('src/propeller/push/instructions/'+file, "r")
# set search text
text = "(def-instruction"
# reading file content line by line.
lines = file_read.readlines()
# looping through each line in the file
# if the line contains "\(def-instruction", go through lines above that line and add
# the Clojure comments to a list which is later written into markdown file.
for count, line in enumerate(lines):
new_list = []
# print(line)
# print(count)
if text in line:
hasDefInstruction = True
# print(line)
mdFile.new_header(level=2, title=lines[count+1].strip())
isComment = True
inc = 1
while isComment:
if lines[count-inc].startswith(';;'):
new_list.append(lines[count-inc].replace(';', '').strip())
# print(lines[count-inc])
inc = inc + 1
else:
isComment = False
new_list.reverse()
for comment in new_list:
mdFile.write(comment+' ')
functionInfo = lines[count+1].strip() + lines[count-1].replace(';', '').strip()
# print(functionInfo)
new_list.append(functionInfo)
# closing file after reading
file_read.close()
# the input string doesn't
# found in the text file
if not hasDefInstruction:
print("\n\"" + text + "\" is not found in \"" + file + "\"!")
mdFile.new_paragraph('')
else:
print("There is"+text)
# entering except block
# if input file doesn't exist
except:
print("\nThe file doesn't exist!")
mdFile.new_table_of_contents()
mdFile.create_md_file()

5
scripts/GenerateDocs.sh Normal file
View File

@ -0,0 +1,5 @@
#!/bin/sh
lein codox
python3 FunctionsToMD.py
python3 HTMLFix.py

27
scripts/HTMLFix.py Normal file
View File

@ -0,0 +1,27 @@
# fixes ordered lists in codox-generated HTML for specific files
import os
os.chdir('..')
htmlFiles = ['Adding_Genetic_Operators.html', 'Adding_Problem.html', 'Adding_Selection_Method.html']
for file in htmlFiles:
with open('docs/'+file, 'r') as f:
OL = "ol>"
countOL = 0
newline = []
for line in f.readlines():
if OL in line:
countOL = countOL + 1
if countOL != 2 and countOL != 3 and countOL != 6 and countOL != 7:
newline.append(line)
else:
newline.append(line)
with open('docs/'+file, 'w') as f:
for line in newline:
f.writelines(line)

View File

@ -0,0 +1,202 @@
# A Guide to Propeller
**Propeller** is an implementation of the Push programming language and the PushGP genetic programming system in Clojure.
For more information on Push and PushGP see http://pushlanguage.org.
## Overview
**Propeller** is a Push-based genetic programming system in Clojure.
<!-- TOC -->
* [A Guide to Propeller](#a-guide-to-propeller)
* [Overview](#overview)
* [What can you do with Propeller?](#what-can-you-do-with-propeller)
* [Installation](#installation)
* [How do I run Propeller on a problem?](#how-do-i-run-propeller-on-a-problem)
* [An Example](#an-example)
* [Can you use a REPL?](#can-you-use-a-repl)
* [Tutorials](#tutorials)
* [Contributing](#contributing)
* [License](#license)
* [Citation](#citation)
* [Contact](#contact)
<!-- TOC -->
### What can you do with Propeller?
You can evolve a Push program to solve a problem.
You can also use the Push interpreter to evaluate Push programs in other projects,
for example in agent-based evolutionary simulations in which
agents are controlled by evolving Push programs.
## Installation
If you have installed [leiningen](https://leiningen.org), which is a tool
for running Clojure programs, then you can run Propeller on a genetic
programming problem that is defined within this project from the command
line with the command `lein run -m <namespace>`, replacing `<namespace>`
with the actual namespace that you will find at the top of the problem file.
If you have installed [Clojure](https://clojure.org/guides/install_clojure#java), you can run Propeller on a genetic programming
problem with the command `clj --main <namespace>`, replacing `<namespace>` with
the actual namespace that you will find at the top of the problem file.
The examples below use leiningen, but you can replace `lein run -m` with `clj --main` to run the same problem.
A specific example is provided later below.
## How do I run Propeller on a problem?
To run Propeller on a problem, you want to call the `-main` function in the problem file using leiningen.
The `-main` function will create a map of arguments from the input and run the main genetic programming loop.
Below is the general format to run a problem through the command-line:
```
lein run -m [namespace of the problem file you want to test]
```
Additional command-line arguments may
be provided to override the default key/value pairs specified in the
problem file,
```
lein run -m [namespace of the problem file you want to test] [key and value] [key and value]...
```
The possible keys come from the table below:
| Key | Description |
|----------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| `:instructions` | List of possible Push instructions used to create a plushy |
| `:error-function` | The error function used to evaluate individuals, specified in the given problem's namespace |
| `:training-data` | Map of inputs and desired outputs used to evaluate individuals of the form: {:input1 first-input :input2 second-input ... :output1 first-output ...} |
| `:testing-data` | Map of inputs and desired outputs not in the training-data to test generalizability of a program that fits the `training-data`. The map is of the form: {:input1 first-input :input2 second-input ... :output1 first-output ...} |
| `:max-generations` | Maximum number of generations |
| `:population-size` | Size of population in a generation |
| `:max-initial-plushy-size` | Maximum number of Push instructions in the initial plushy |
| `:step-limit` | The maximum number of steps that a Push program will be executed by `interpret-program` |
| `:parent-selection` | Function from `propeller.selection` that determines method of parent selection method. Propeller includes `:tournament-selection`, `:lexicase-selection`, and `:epsilon-lexicase-selection`. |
| `:tournament-size` | If using a tournament selection method, the number of individuals in each tournaments used to determine parents |
| `:umad-rate` | Rate (decimal between 0 and 1) of uniform mutation by addition and deletion (UMAD) genetic operator |
| `:variation` | Map with genetic operators as keys and probabilities as values. For example, {:umad 0.3 :crossover 0.7}. This would mean that when the system needs to generate a child, it will use UMAD 30% of the time and crossover 70% of the time. The probabilities should sum to 1. |
| `:elitism` | When true, will cause the individual with the lowest error in the population to survive, without variation, into the next generation. |
When you run a problem, you will get a report each generation with the following information:
```
:generation
:best-plushy
:best-program
:best-total-error
:best-errors
:best-behaviors
:genotypic-diversity
:behavioral-diversity
:average-genome-length
:average-total-error
```
### An Example
For example, you can run the simple-regression genetic programming problem with:
```
lein run -m propeller.problems.simple-regression
```
This will run simple-regression with the default set of arguments in the `simple-regression` problem file.
```
{:instructions instructions
:error-function error-function
:training-data (:train train-and-test-data)
:testing-data (:test train-and-test-data)
:max-generations 500
:population-size 500
:max-initial-plushy-size 100
:step-limit 200
:parent-selection :lexicase
:tournament-size 5
:umad-rate 0.1
:variation {:umad 0.5 :crossover 0.5}
:elitism false}
```
You can override the default key/value pairs with additional arguments. For example:
```
lein run -m propeller.problems.simple-regression :population-size 100
```
On Unix operating systems, including MacOS, you can use something
like the following to send output both to the terminal
and to a text file (called `outfile` in this example):
```
lein run -m propeller.problems.simple-regression | tee outfile
```
If you want to provide command line arguments that include
characters that may be interpreted by your command line shell
before they get to Clojure, then enclose those in double
quotes, like in this example that provides a non-default
value for the `:variation` argument, which is a clojure map
containing curly brackets that may confuse your shell:
```
lein run -m propeller.problems.simple-regression :variation "{:umad 1.0}"
```
### Can you use a REPL?
Yes!
To run a genetic programming problem from a REPL, start
your REPL for the project (e.g. with `lein repl` at the
command line when in the project directory, or through your
IDE) and then do something like the following (which in
this case runs the simple-regression problem with
`:population-size` 100):
```
(require 'propeller.problems.simple-regression)
(in-ns 'propeller.problems.simple-regression)
(-main :population-size 100 :variation {:umad 1.0})
```
If you want to run the problem with the default parameters,
then you should call `-main` without arguments, as `(-main).
## Tutorials
- [Adding genetic operators](Adding_Genetic_Operators.md)
- [Adding selection methods](Adding_Selection_Method.md)
- [Adding a new problem](Adding_Problem.md)
- [Generating Documentation](Generating_Documentation.md)
## Contributing
You can report a bug on the [GitHub issues page](https://github.com/lspector/propeller/issues).
The best way to contribute to Propeller is to fork the [main GitHub repository](https://github.com/lspector/propeller) and submit a pull request.
Propeller provides a way to automatically [generate documentation](Generating_Documentation.md) for any contributions
you might make.
## License
**Eclipse Public License 2.0**
This commercially-friendly copyleft license provides the ability to commercially license binaries;
a modern royalty-free patent license grant; and the ability for linked works to use other licenses, including commercial ones.
## Citation
We are in the process of creating a DOI, but in the meantime,
we ask that you cite the [link to the repository](https://github.com/lspector/propeller) if you use Propeller.
## Contact
To discuss Propeller, Push, and PushGP, you can join the [Push-Language Discourse](https://discourse.pushlanguage.org/).

View File

@ -0,0 +1,62 @@
# Adding Genetic Operators
In addition to the already-included genetic operators, you can add your own!
## Variation Genetic Operators
1. Go to `propeller.variation.cljc`
2. Define a genetic operator function
3. In `propeller.variation/new-individual`, add the new genetic operator in the `new-individual` function under the `case` call
``` clojure
(defn new-individual
"Returns a new individual produced by selection and variation of
individuals in the population."
[pop argmap]
...
(case op
...
:new-genetic-operator
(-> (:plushy (selection/select-parent pop argmap))
(new-genetic-operator ))
...
:else
(throw #?(:clj (Exception. (str "No match in new-individual for " op))
:cljs (js/Error
(str "No match in new-individual for " op))))))})
```
4. When running a problem, specify the genetic operator in `:variation`.
For example:
```
lein run -m propeller.problems.simple-regression :variation "{:new-genetic-operator 1.0}"
```
## Selection Genetic Operators
1. Go to `propeller.selection.cljc`
2. Define a genetic operator function
3. In `propeller.selection.cljc`, add the new genetic operator in the `select-parent` function under the `case` call.
```clojure
(defn select-parent
"Selects a parent from the population using the specified method."
[pop argmap]
(case (:parent-selection argmap)
...
:new-genetic-operator (:new-genetic-operator )
...
))
```
4. When running a problem, specify the selection method in `:parent-selection`
For example:
```
lein run -m propeller.problems.simple-regression :parent-selection :new-genetic-operator
```

View File

@ -0,0 +1,122 @@
# Adding a Problem
In general, a problem file has 3 components: `train-and-test-data`, `instructions`, `error-function`, and `-main`.
1. To add a new problem, you need training and test data. For Problem Synthesis Benchmark Problems (PSB2),
you can fetch datasets using `psb2.core/fetch-examples`.
```clojure
(defn fetch-examples
"Fetches and returns training and test data from a PSB2 problem.
Returns a map of the form {:train training-examples :test testing-examples}
where training-examples and testing-examples are lists of training and test
data. The elements of these lists are maps of the form:
{:input1 first-input :input2 second-input ... :output1 first-output ...}
The training examples will include all hard-coded edge cases included in the suite,
along with enough random cases to include `n-train` cases.
Note that this function loads large datasets and can be slow, 30-120 seconds.
Parameters:
`datasets-directory` - Location of the PSB2 datasets as downloaded from https://zenodo.org/record/4678739
`problem-name` - Name of the PSB2 problem, lowercase and seperated by dashes.
- Ex: indices-of-substring
`n-train` - Number of training cases to return
`n-test` - Number of test cases to return"
[datasets-directory problem-name n-train n-test]
```
2. Define the possible Push instructions to be used to create plushys. It should be a non-lazy list of
instructions from `push/instructions`, input instructions, close, and constants (including functions that produce constants).
3. Define an error function that will evaluate plushys and add `:behaviors parsed-outputs`,
`:errors`, and `:total-error` to the individual
4. Define the function `-main` with a map of default arguments.
## Example of a Problem
```clojure
(ns propeller.problems.PSB2.solve-boolean
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
[propeller.utils :as utils]
[propeller.push.instructions :refer [get-stack-instructions]]
[propeller.push.state :as state]
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ================================
; SOLVE BOOLEAN from PSB2
; Given a string representing a Boolean
; expression consisting of T, F, |, and &, evaluate it and return
; the resulting Boolean.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ==================================================================
(def train-and-test-data (psb2/fetch-examples "data" "solve-boolean" 200 2000))
(def instructions
(utils/not-lazy
(concat
;;; stack-specific instructions
(get-stack-instructions #{:exec :integer :boolean :char :string :print})
;;; input instructions
(list :in1)
;;; close
(list 'close)
;;; ERCs (constants)
(list true false \t \f \& \|))))
(defn error-function
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
correct-outputs (map (fn [i] (get i :output1)) data)
outputs (map (fn [input]
(state/peek-stack
(interpreter/interpret-program
program
(assoc state/empty-state :input {:in1 input})
(:step-limit argmap))
:boolean))
inputs)
parsed-outputs (map (fn [output]
(try (read-string output)
#?(:clj (catch Exception e 1000.0)
:cljs (catch js/Error. e 1000.0))))
outputs)
errors (map (fn [correct-output output]
(if (= output :no-stack-item)
10000
(if (= correct-output output)
0
1)))
correct-outputs
parsed-outputs)]
(assoc individual
:behaviors parsed-outputs
:errors errors
:total-error #?(:clj (apply +' errors)
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
[& args]
(gp/gp
(merge
{:instructions instructions
:error-function error-function
:training-data (:train train-and-test-data)
:testing-data (:test train-and-test-data)
:max-generations 300
:population-size 1000
:max-initial-plushy-size 250
:step-limit 2000
:parent-selection :lexicase
:tournament-size 5
:umad-rate 0.1
:variation {:umad 1.0 :crossover 0.0}
:elitism false}
(apply hash-map (map #(if (string? %) (read-string %) %) args))))
(#?(:clj shutdown-agents)))
```

View File

@ -0,0 +1,20 @@
# Adding a Selection Method
1. Define a selection method function in `propeller.selection` that selects an individual from the population
2. Add the selection method in `propeller.selection/select-parent` under the `case` call:
```clojure
(defn select-parent
"Selects a parent from the population using the specified method."
[pop argmap]
(case (:parent-selection argmap)
:new-selection-method (new-selection-method )))
```
3. When runnning a problem, specify the selection method in `:parent-selection`.
For example:
```
lein run -m propeller.problems.simple-regression :parent-selection :new-selection-method"
```

View File

@ -0,0 +1,151 @@
Table of contents
=================
* [Additional Instructions](#additional-instructions)
* [input_output.cljc](#input_outputcljc)
* [numeric.cljc](#numericcljc)
* [string.cljc](#stringcljc)
* [character.cljc](#charactercljc)
* [bool.cljc](#boolcljc)
* [code.cljc](#codecljc)
# Additional Instructions
# input_output.cljc
## :print_newline
Prints new line
# numeric.cljc
## :float_cos
Pushes the cosine of the top FLOAT
## :float_sin
Pushes the sine of the top FLOAT
## :float_tan
Pushes the tangent of the top FLOAT
## :float_from_integer
Pushes the floating point version of the top INTEGER
## :integer_from_float
Pushes the result of truncating the top FLOAT towards negative infinity
# string.cljc
## :string_butlast
Pushes the butlast of the top STRING (i.e. the string without its last letter)
## :string_concat
Pushes the concatenation of the top two STRINGs (second + first)
## :string_conj_char
Pushes the concatenation of the top STRING and the top CHAR (STRING + CHAR)
## :string_contains
Pushes TRUE if the top STRING is a substring of the second STRING, and FALSE otherwise
## :string_contains_char
Pushes TRUE if the top CHAR is contained in the top STRING, and FALSE otherwise
## :string_drop
Pushes the top STRING with n characters dropped, where n is taken from the top of the INTEGER stack
## :string_empty_string
Pushes TRUE if the top STRING is the empty string
## :string_first
Pushes the first CHAR of the top STRING
## :string_from_boolean
Pushes the STRING version of the top BOOLEAN, e.g. "true"
## :string_from_char
Pushes the STRING version of the top CHAR, e.g. "a"
## :string_from_float
Pushes the STRING version of the top FLOAT e.g. "2.05"
## :string_from_integer
Pushes the STRING version of the top INTEGER, e.g. "3"
## :string_indexof_char
Pushes the index of the top CHAR in the top STRING onto the INTEGER stack. If the top CHAR is not present in the top string, acts as a NOOP
## :string_iterate
Iterates over the top STRING using code on the EXEC stack
## :string_last
Pushes the last CHAR of the top STRING. If the string is empty, do nothing
## :string_length
Pushes the length of the top STRING onto the INTEGER stack
## :string_nth
Pushes the nth CHAR of the top STRING, where n is taken from the top of the INTEGER stack. If n exceeds the length of the string, it is reduced modulo the length of the string
## :string_occurencesof_char
Pushes the number of times the top CHAR occurs in the top STRING onto the INTEGER stack
## :string_parse_to_chars
Splits the top string into substrings of length 1 (i.e. into its component characters) and pushes them back onto the STRING stack in the same order
## :string_remove_char
Pushes the top STRING, with all occurrences of the top CHAR removed
## :string_replace
Pushes the third topmost STRING on stack, with all occurences of the second topmost STRING replaced by the top STRING
## :string_replace_char
Pushes the top STRING, with all occurences of the second topmost CHAR replaced with the top CHAR
## :string_replace_first
Pushes the third topmost STRING on stack, with the first occurence of the second topmost STRING replaced by the top STRING
## :string_replace_first_char
Pushes the top STRING, with the first occurence of the second topmost CHAR replaced with the top CHAR
## :string_rest
Pushes the rest of the top STRING (i.e. the string without its first letter)
## :string_reverse
Pushes the reverse of the top STRING
## :string_set_char
Pushes the top STRING, with the letter at index n (where n is taken from the INTEGER stack) replaced with the top CHAR. If n is out of bounds, it is reduced modulo the length of the string
## :string_split
Splits the top STRING on whitespace, and pushes back the resulting components in the same order
## :string_substr
Pushes the substring of the top STRING, with beginning and end indices determined by the second topmost and topmost INTEGERs respectively. If an index is out of bounds, the beginning/end of the string is used instead
## :string_take
Pushes the substring of the top STRING consisting of its first n letters, where n is determined by the top INTEGER
# character.cljc
## :char_is_letter
Pushes TRUE onto the BOOLEAN stack if the popped character is a letter
## :char_is_digit
Pushes TRUE onto the BOOLEAN stack if the popped character is a digit
## :char_is_whitespace
Pushes TRUE onto the BOOLEAN stack if the popped character is whitespace (newline, space, or tab)
## :char_from_float
Pops the FLOAT stack, converts the top item to a whole number, and pushes its corresponding ASCII value onto the CHAR stack. Whole numbers larger than 128 will be reduced modulo 128. For instance, 248.45 will result in x being pushed.
## :char_from_integer
Pops the INTEGER stack and pushes the top element's corresponding ASCII value onto the CHAR stack. Integers larger than 128 will be reduced modulo 128. For instance, 248 will result in x being pushed
## :char_all_from_string
Pops the STRING stack and pushes the top element's constituent characters onto the CHAR stack, in order. For instance, "hello" will result in the top of the CHAR stack being \h \e \l \l \o
# bool.cljc
## :boolean_and
Pushes the logical AND of the top two BOOLEANs
## :boolean_or
Pushes the logical OR of the top two BOOLEANs
## :boolean_not
Pushes the logical NOT of the top BOOLEAN
## :boolean_xor
Pushes the logical XOR of the top two BOOLEAN
## :boolean_invert_first_then_and
Pushes the logical AND of the top two BOOLEANs, after applying NOT to the first one
## :boolean_invert_second_then_and
Pushes the logical AND of the top two BOOLEANs, after applying NOT to the second one
## :boolean_from_float
Pushes FALSE if the top FLOAT is 0.0, and TRUE otherwise
## :boolean_from_integer
Pushes FALSE if the top INTEGER is 0, and TRUE otherwise
# code.cljc
## :code_append
Concatenates the top two instructions on the :code stack and pushes the result back onto the stack
## :exec_do_range
Executes the top EXEC instruction (i.e. loops) a number of times determined by the top two INTEGERs, while also pushing the loop counter onto the INTEGER stack. The top INTEGER is the "destination index" and the second INTEGER is the "current index". If the integers are equal, then the current index is pushed onto the INTEGER stack and the code (which is the "body" of the loop) is pushed onto the EXEC stack for subsequent execution. If the integers are not equal, then the current index will still be pushed onto the INTEGER stack but two items will be pushed onto the EXEC stack - first a recursive call to :exec_do_range (with the same code and destination index, but with a current index that has been either incremented or decremented by 1 to be closer to the destination index) and then the body code. Note that the range is inclusive of both endpoints a call with integer arguments 3 and 5 will cause its body to be executed 3 times, with the loop counter having the values 3, 4, and 5. Note also that one can specify a loop that "counts down" by providing a destination index that is less than the specified current index.
## :exec_do_count
Executes the top EXEC instruction (i.e. loops) a number of times determined by the top INTEGER, pushing an index (which runs from 0 to one less than the total number of iterations) onto the INTEGER stack prior to each execution of the loop body. If the top INTEGER argument is <= 0, this becomes a NOOP
## :exec_do_times
Like :exec_do_count, but does not push the loop counter onto the INTEGER stack
## :exec_if
If the top BOOLEAN is TRUE, removes the the second item on the EXEC stack, leaving the first item to be executed. Otherwise, removes the first item, leaving the second to be executed. Acts as a NOOP unless there are at least two items on the EXEC stack and one item on the BOOLEAN stack
## :exec_when
If the top BOOLEAN is TRUE, leaves the first item on the EXEC stack to be executed. Otherwise, it removes it. Acts as a NOOP unless there is at least one item on the EXEC stack and one item on the BOOLEAN stack
## :exec_while
Keeps executing the top instruction on the EXEC stack while the top item on the BOOLEAN stack is true
## :exec_do_while
Keeps executing the top instruction on the EXEC stack while the top item on the BOOLEAN stack is true. Differs from :exec_while in that it executes the top instruction at least once
## :exec_k
The "K combinator" - removes the second item on the EXEC stack
## :exec_s
The "S combinator" - pops 3 items from the EXEC stack, which we will call A, B, and C (with A being the first one popped), and then pushes a list containing B and C back onto the EXEC stack, followed by another instance of C, followed by another instance of A
## :exec_y
The "Y combinator" - inserts beneath the top item of the EXEC stack a new item of the form "(:exec_y TOP_ITEM)"

View File

@ -0,0 +1,8 @@
# Generating Documentation for Propeller
To generate documentation with [codox](https://github.com/weavejester/codox), run `scripts/GenerateDocs.sh`in the command line.
This will run "lein codox" on the command line to generate first batch of HTMl files.
Then, it runs FunctionsToMD to take Push instructions generated by `def-instruction` and spit it out to a Markdown file.
Then, it runs HTMLFix to fix the ordered lists in `Adding_Genetic_Operators.md`, `Adding_Problem.md`, and
`Adding_Selection_Method.md`.

View File

@ -1,4 +1,4 @@
(ns propeller.core
(ns ^:no-doc propeller.core
#?(:clj (:gen-class)))
(defn -main
@ -7,4 +7,4 @@
;; Exception for when no args were passed
(println "To run a genetic programming problem, provide a the problem's")
(println "namespace as specified in the Propeller README file at")
(println "https://github.com/lspector/propeller/blob/master/README.md"))
(println "https://github.com/lspector/propeller/blob/master/A_Guide_To_Propeller.md"))

View File

@ -1,4 +1,7 @@
(ns propeller.genome
"The genetic material in Propeller. A `plushy` is a list of Push instructions that represent a Push program.
They hold the genetic material for an `individual`. In the initial population, we create random plushys."
{:doc/format :markdown}
(:require [propeller.push.instructions :as instructions]
[propeller.utils :as utils]))
@ -10,7 +13,12 @@
#(utils/random-instruction instructions)))
(defn plushy->push
"Returns the Push program expressed by the given plushy representation."
"Returns the Push program expressed by the given plushy representation.
The function takes in a plushy representation as input and converts it into a Push program by iteratively processing
the plushy elements and adding instructions to the push program.
It also handles the case where there are open instructions that need to be closed before the end of the program.
"
([plushy] (plushy->push plushy {}))
([plushy argmap]
(let [plushy (if (:diploid argmap) (map first (partition 2 plushy)) plushy)

View File

@ -1,4 +1,5 @@
(ns propeller.gp
"Main genetic programming loop."
(:require [clojure.string]
[clojure.pprint]
[propeller.genome :as genome]
@ -35,7 +36,25 @@
(println)))
(defn gp
"Main GP loop."
"Main GP loop.
On each iteration, it creates a population of random plushies using a mapper
function and genome/make-random-plushy function,
then it sorts the population by the total error using the error-function
and sort-by function. It then takes the best individual from the sorted population,
and if the parent selection is set to epsilon-lexicase, it adds the epsilons to the argmap.
The function then checks if the custom-report argument is set,
if so it calls that function passing the evaluated population,
current generation and argmap. If not, it calls the report function
passing the evaluated population, current generation and argmap.
Then, it checks if the total error of the best individual is less than or equal
to the solution-error-threshold or if the current generation is greater than or
equal to the max-generations specified. If either is true, the function
exits with the best individual or nil. If not, it creates new individuals
for the next generation using the variation/new-individual function and the
repeatedly function, and then continues to the next iteration of the loop. "
[{:keys [population-size max-generations error-function instructions
max-initial-plushy-size solution-error-threshold mapper ds-parent-rate ds-parent-gens dont-end ids-type downsample?]
:or {solution-error-threshold 0.0

View File

@ -1,4 +1,4 @@
(ns propeller.main
(ns ^:no-doc propeller.main
(:require [propeller.core :as propeller]))
(defn main! []

View File

@ -56,7 +56,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -74,7 +74,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -125,7 +125,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -57,7 +57,9 @@
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,12 @@
(ns propeller.problems.PSB2.basement
"BASEMENT from PSB2
Given a vector of integers, return the first
index such that the sum of all integers from the start of the
vector to that index (inclusive) is negative.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,21 +17,14 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ============================
; BASEMENT from PSB2
; Given a vector of integers, return the first
; index such that the sum of all integers from the start of the
; vector to that index (inclusive) is negative.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ===============================================================
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "basement" 200 2000))
(def train-and-test-data (psb2/fetch-examples "data" "basement" 200 2000))
; Random integer between -100 and 100 (from smallest)
(defn random-int [] (- (rand-int 201) 100))
(defn random-int
"Random integer between -100 and 100 (from smallest)"
[] (- (rand-int 201) 100))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -36,6 +37,10 @@
(list random-int -1 0 1 []))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the INTEGER stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -61,7 +66,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,14 @@
(ns propeller.problems.PSB2.bouncing-balls
"BOUNCING BALLS from PSB2
Given a starting height and a height after the first bounce of a
dropped ball, calculate the bounciness index
(height of first bounce / starting height). Then, given a number
of bounces, use the bounciness index to calculate the total
distance that the ball travels across those bounces.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,18 +19,8 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ===============================
; BOUNCING BALLS from PSB2
; Given a starting height and a height after the first bounce of a
; dropped ball, calculate the bounciness index
; (height of first bounce / starting height). Then, given a number
; of bounces, use the bounciness index to calculate the total
; distance that the ball travels across those bounces.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ==================================================================
(def train-and-test-data (psb2/fetch-examples "data" "bouncing-balls" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "bouncing-balls" 200 2000))
(defn map-vals-input
"Returns all the input values of a map (specific helper method for bouncing-balls)"
@ -33,6 +33,7 @@
(get i :output1))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -45,6 +46,10 @@
(list 0.0 1.0 2.0))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the FLOAT stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (map-vals-input i)) data)
@ -72,7 +77,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,12 @@
(ns propeller.problems.PSB2.bowling
"BOWLING from PSB2
Given a string representing the individual
bowls in a 10-frame round of 10 pin bowling, return the
score of that round.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,20 +17,14 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ======================
; BOWLING from PSB2
; Given a string representing the individual
; bowls in a 10-frame round of 10 pin bowling, return the
; score of that round.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; =========================================================
(def train-and-test-data (psb2/fetch-examples "data" "bowling" 200 2000))
(defn random-int [] (- (rand-int 201) 100))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "bowling" 200 2000))
(defn random-int "Returns random integer between -100 and 100" [] (- (rand-int 201) 100))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -35,6 +37,10 @@
(list \- \X \/ \1 \2 \3 \4 \5 \6 \7 \8 \9 10 random-int))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the INTEGER stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -60,7 +66,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,13 @@
(ns propeller.problems.PSB2.camel-case
"CAMEL CASE from PSB2
Take a string in kebab-case and convert all of the words to camelCase.
Each group of words to convert is delimited by \"-\", and each grouping
is separated by a space. For example: \"camel-case example-test-string\"
→ \"camelCase exampleTestString\"
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,25 +18,18 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION =====================================
; CAMEL CASE from PSB2
; Take a string in kebab-case and convert all of the words to camelCase.
; Each group of words to convert is delimited by "-", and each grouping
; is separated by a space. For example: "camel-case example-test-string"
; → "camelCase exampleTestString"
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; =======================================================================
(def train-and-test-data (psb2/fetch-examples "data" "camel-case" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "camel-case" 200 2000))
; Visible character ERC
(defn random-char
"Return visible character ERC"
[]
(rand-nth (map char (range 97 122))))
; Word generator for string ERC
(defn word-generator
"Word generator for string ERC"
[]
(let [chars-between #(map char (range (int %1) (inc (int %2))))
chars (chars-between \a \z)
@ -35,6 +37,7 @@
(apply str (repeatedly word-len #(rand-nth chars)))))
(defn cleanup-length
"Remove spaces and dashes from end of string"
[string len]
(let [result (take len string)]
(if (or (= (last result) \space)
@ -44,6 +47,7 @@
; String ERC
(defn random-input
"Returns random string ERCs"
[len]
(loop [result-string (word-generator)]
(if (>= (count result-string) len)
@ -53,6 +57,7 @@
(word-generator))))))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -66,6 +71,10 @@
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the STRING stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -91,7 +100,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,12 @@
(ns propeller.problems.PSB2.dice-game
"DICE GAME from PSB2
Peter has an n sided die and Colin has an m
sided die. If they both roll their dice at the same time, return
the probability that Peter rolls strictly higher than Colin.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,16 +17,8 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ===============================
; DICE GAME from PSB2
; Peter has an n sided die and Colin has an m
; sided die. If they both roll their dice at the same time, return
; the probability that Peter rolls strictly higher than Colin.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ==================================================================
(def train-and-test-data (psb2/fetch-examples "data" "dice-game" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "dice-game" 200 2000))
(defn map-vals-input
"Returns all the input values of a map"
@ -31,6 +31,7 @@
(get i :output1))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -43,6 +44,10 @@
(list 0.0 1.0))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the FLOAT stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (map-vals-input i)) data)
@ -69,7 +74,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,12 @@
(ns propeller.problems.PSB2.fizz-buzz
"FIZZ BUZZ from PSB2
Given an integer x, return \"Fizz\" if x is
divisible by 3, \"Buzz\" if x is divisible by 5, \"FizzBuzz\" if x
is divisible by 3 and 5, and a string version of x if none of
the above hold.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -20,11 +28,14 @@
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ============================================================
(def train-and-test-data (psb2/fetch-examples "data" "fizz-buzz" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "fizz-buzz" 200 2000))
(def train-data (:train train-and-test-data))
(def test-data (:test train-and-test-data))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -37,6 +48,10 @@
(list "Fizz" "Buzz" "FizzBuzz" 0 3 5))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the STRING stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -63,7 +78,9 @@
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,13 @@
(ns propeller.problems.PSB2.fuel-cost
"FUEL COST from PSB2
Given a vector of positive integers, divide
each by 3, round the result down to the nearest integer, and
subtract 2. Return the sum of all of the new integers in the
vector
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -20,14 +29,15 @@
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ============================================================
(def train-and-test-data (psb2/fetch-examples "data" "fuel-cost" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "fuel-cost" 200 2000))
(def train-data (:train train-and-test-data))
(def test-data (:test train-and-test-data))
; Random integer between -100 and 100 (from smallest)
(defn random-int [] (- (rand-int 201) 100))
(defn random-int "Random integer between -100 and 100" [] (- (rand-int 201) 100))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -40,6 +50,10 @@
(list random-int 0 1 2 3))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the INTEGER stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -65,7 +79,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,11 @@
(ns propeller.problems.PSB2.gcd
"GCD [GREATEST COMMON DIVISOR] from PSB2
Given two integers, return the largest integer that divides each
of the integers evenly
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -10,20 +17,14 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ===============================
; GCD [GREATEST COMMON DIVISOR] from PSB2
; Given two integers, return the largest integer that divides each
; of the integers evenly
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ==================================================================
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "gcd" 200 2000))
(def train-and-test-data (psb2/fetch-examples "data" "gcd" 200 2000))
(def train-data (:train train-and-test-data))
(def test-data (:test train-and-test-data))
(defn random-int [] (- (rand-int 201) 100))
(defn random-int "Random integer between -100 and 100" [] (- (rand-int 201) 100))
(defn map-vals-input
"Returns all the input values of a map"
@ -36,6 +37,7 @@
(get i :output1))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -48,6 +50,10 @@
(list random-int))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the INTEGER stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (map-vals-input i)) data)
@ -74,7 +80,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,14 @@
(ns propeller.problems.PSB2.luhn
"LUHN from PSB2
Given a vector of 16 digits, implement Luhns
algorithm to verify a credit card number, such that it follows
the following rules: double every other digit starting with
the second digit. If any of the results are over 9, subtract 9
from them. Return the sum of all of the new digits.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,23 +19,14 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION ============================
; LUHN from PSB2
; Given a vector of 16 digits, implement Luhns
; algorithm to verify a credit card number, such that it follows
; the following rules: double every other digit starting with
; the second digit. If any of the results are over 9, subtract 9
; from them. Return the sum of all of the new digits.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ===============================================================
(def train-and-test-data (psb2/fetch-examples "data" "luhn" 200 2000))
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "luhn" 200 2000))
; Random integer between -100 and 100 (from smallest)
(defn random-int [] (- (rand-int 201) 100))
(defn random-int "Random integer between -100 and 100" [] (- (rand-int 201) 100))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -38,6 +39,10 @@
(list 0 2 9 10 random-int))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the INTEGER stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -63,7 +68,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

View File

@ -1,4 +1,12 @@
(ns propeller.problems.PSB2.middle-character
"MIDDLE CHARACTER from PSB2
Given a string, return the middle
character as a string if it is odd length; return the two middle
characters as a string if it is even length.
Source: https://arxiv.org/pdf/2106.06086.pdf"
{:doc/format :markdown}
(:require [psb2.core :as psb2]
[propeller.genome :as genome]
[propeller.push.interpreter :as interpreter]
@ -9,20 +17,12 @@
[propeller.gp :as gp]
#?(:cljs [cljs.reader :refer [read-string]])))
; =========== PROBLEM DESCRIPTION =============================
; MIDDLE CHARACTER from PSB2
; Given a string, return the middle
; character as a string if it is odd length; return the two middle
; characters as a string if it is even length.
;
; Source: https://arxiv.org/pdf/2106.06086.pdf
; ===============================================================
(def train-and-test-data "Data taken from https://zenodo.org/record/5084812" (psb2/fetch-examples "data" "middle-character" 200 2000))
(def train-and-test-data (psb2/fetch-examples "data" "middle-character" 200 2000))
(defn random-int [] (- (rand-int 201) 100))
(defn random-int "Random integer between -100 and 100" [] (- (rand-int 201) 100))
(def instructions
"Stack-specific instructions, input instructions, close, and constants"
(utils/not-lazy
(concat
;;; stack-specific instructions
@ -35,6 +35,10 @@
(list "" 0 1 2 random-int))))
(defn error-function
"Finds the behaviors and errors of an individual: Error is 0 if the value and
the program's selected behavior match, or 1 if they differ, or 1000000 if no
behavior is produced. The behavior is here defined as the final top item on
the STRING stack."
[argmap data individual]
(let [program (genome/plushy->push (:plushy individual) argmap)
inputs (map (fn [i] (get i :input1)) data)
@ -60,7 +64,9 @@
:cljs (apply + errors)))))
(defn -main
"Runs propel-gp, giving it a map of arguments."
"Runs the top-level genetic programming function, giving it a map of
arguments with defaults that can be overridden from the command line
or through a passed map."
[& args]
(gp/gp
(merge

Some files were not shown because too many files have changed in this diff Show More