galaaz 0.4.6 → 0.5.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +5 -5
- data/README.md +3575 -118
- data/Rakefile +21 -4
- data/bin/gknit +152 -6
- data/bin/gknit-draft +105 -0
- data/bin/gknit-draft.rb +28 -0
- data/bin/gknit_Rscript +127 -0
- data/bin/grun +27 -1
- data/bin/gstudio +47 -4
- data/bin/{gstudio.rb → gstudio_irb.rb} +0 -0
- data/bin/gstudio_pry.rb +7 -0
- data/blogs/galaaz_ggplot/galaaz_ggplot.Rmd +3 -12
- data/blogs/galaaz_ggplot/galaaz_ggplot.html +77 -222
- data/blogs/galaaz_ggplot/galaaz_ggplot.md +4 -31
- data/blogs/galaaz_ggplot/galaaz_ggplot.pdf +0 -0
- data/blogs/galaaz_ggplot/galaaz_ggplot_files/figure-html/midwest_rb.png +0 -0
- data/blogs/galaaz_ggplot/galaaz_ggplot_files/figure-html/scatter_plot_rb.png +0 -0
- data/blogs/galaaz_ggplot/midwest.Rmd +1 -9
- data/blogs/gknit/gknit.Rmd +232 -123
- data/blogs/{dev/dev.html → gknit/gknit.html} +1897 -33
- data/blogs/gknit/gknit.pdf +0 -0
- data/blogs/gknit/lst.rds +0 -0
- data/blogs/gknit/stats.bib +27 -0
- data/blogs/manual/lst.rds +0 -0
- data/blogs/manual/manual.Rmd +1893 -47
- data/blogs/manual/manual.html +3153 -347
- data/blogs/manual/manual.md +3575 -118
- data/blogs/manual/manual.pdf +0 -0
- data/blogs/manual/manual.tex +4026 -0
- data/blogs/manual/manual_files/figure-html/bubble-1.png +0 -0
- data/blogs/manual/manual_files/figure-html/diverging_bar.png +0 -0
- data/blogs/manual/manual_files/figure-latex/bubble-1.png +0 -0
- data/blogs/manual/manual_files/figure-latex/diverging_bar.pdf +0 -0
- data/blogs/{dev → manual}/model.rb +0 -0
- data/blogs/nse_dplyr/nse_dplyr.Rmd +849 -0
- data/blogs/nse_dplyr/nse_dplyr.html +878 -0
- data/blogs/nse_dplyr/nse_dplyr.md +1198 -0
- data/blogs/nse_dplyr/nse_dplyr.pdf +0 -0
- data/blogs/oh_my/oh_my.html +274 -386
- data/blogs/oh_my/oh_my.md +208 -205
- data/blogs/ruby_plot/ruby_plot.Rmd +64 -84
- data/blogs/ruby_plot/ruby_plot.html +235 -208
- data/blogs/ruby_plot/ruby_plot.md +239 -34
- data/blogs/ruby_plot/ruby_plot.pdf +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/dose_len.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facet_by_delivery.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facet_by_dose.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_by_delivery_color.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_by_delivery_color2.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_with_decorations.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_with_jitter.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_with_points.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/final_box_plot.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/final_violin_plot.png +0 -0
- data/blogs/ruby_plot/ruby_plot_files/figure-html/violin_with_jitter.png +0 -0
- data/examples/Bibliography/master.bib +50 -0
- data/examples/Bibliography/stats.bib +72 -0
- data/examples/islr/ch2.spec.rb +1 -1
- data/examples/islr/ch3_boston.rb +4 -4
- data/examples/islr/x_y_rnorm.jpg +0 -0
- data/examples/latex_templates/Test-acm_article/Makefile +16 -0
- data/examples/latex_templates/Test-acm_article/Test-acm_article.Rmd +65 -0
- data/examples/latex_templates/Test-acm_article/acm_proc_article-sp.cls +1670 -0
- data/examples/latex_templates/Test-acm_article/sensys-abstract.cls +703 -0
- data/examples/latex_templates/Test-acm_article/sigproc.bib +59 -0
- data/examples/latex_templates/Test-acs_article/Test-acs_article.Rmd +260 -0
- data/examples/latex_templates/Test-acs_article/Test-acs_article.pdf +0 -0
- data/examples/latex_templates/Test-acs_article/acs-Test-acs_article.bib +11 -0
- data/examples/latex_templates/Test-acs_article/acs-my_output.bib +11 -0
- data/examples/latex_templates/Test-acs_article/acstest.bib +17 -0
- data/examples/latex_templates/Test-aea_article/AEA.cls +1414 -0
- data/examples/latex_templates/Test-aea_article/BibFile.bib +0 -0
- data/examples/latex_templates/Test-aea_article/Test-aea_article.Rmd +108 -0
- data/examples/latex_templates/Test-aea_article/Test-aea_article.pdf +0 -0
- data/examples/latex_templates/Test-aea_article/aea.bst +1269 -0
- data/examples/latex_templates/Test-aea_article/multicol.sty +853 -0
- data/examples/latex_templates/Test-aea_article/references.bib +0 -0
- data/examples/latex_templates/Test-aea_article/setspace.sty +546 -0
- data/examples/latex_templates/Test-amq_article/Test-amq_article.Rmd +256 -0
- data/examples/latex_templates/Test-amq_article/Test-amq_article.pdf +0 -0
- data/examples/latex_templates/Test-amq_article/Test-amq_article.pdfsync +3397 -0
- data/examples/latex_templates/Test-amq_article/pics/Figure2.pdf +0 -0
- data/examples/latex_templates/Test-ams_article/Test-ams_article.Rmd +215 -0
- data/examples/latex_templates/Test-ams_article/amstest.bib +436 -0
- data/examples/latex_templates/Test-asa_article/Test-asa_article.Rmd +153 -0
- data/examples/latex_templates/Test-asa_article/Test-asa_article.pdf +0 -0
- data/examples/latex_templates/Test-asa_article/agsm.bst +1353 -0
- data/examples/latex_templates/Test-asa_article/bibliography.bib +233 -0
- data/examples/latex_templates/Test-ieee_article/IEEEtran.bst +2409 -0
- data/examples/latex_templates/Test-ieee_article/IEEEtran.cls +6346 -0
- data/examples/latex_templates/Test-ieee_article/Test-ieee_article.Rmd +175 -0
- data/examples/latex_templates/Test-ieee_article/Test-ieee_article.pdf +0 -0
- data/examples/latex_templates/Test-ieee_article/mybibfile.bib +20 -0
- data/examples/latex_templates/Test-rjournal_article/RJournal.sty +335 -0
- data/examples/latex_templates/Test-rjournal_article/RJreferences.bib +18 -0
- data/examples/latex_templates/Test-rjournal_article/RJwrapper.pdf +0 -0
- data/examples/latex_templates/Test-rjournal_article/Test-rjournal_article.Rmd +52 -0
- data/examples/latex_templates/Test-springer_article/Test-springer_article.Rmd +65 -0
- data/examples/latex_templates/Test-springer_article/Test-springer_article.pdf +0 -0
- data/examples/latex_templates/Test-springer_article/bibliography.bib +26 -0
- data/examples/latex_templates/Test-springer_article/spbasic.bst +1658 -0
- data/examples/latex_templates/Test-springer_article/spmpsci.bst +1512 -0
- data/examples/latex_templates/Test-springer_article/spphys.bst +1443 -0
- data/examples/latex_templates/Test-springer_article/svglov3.clo +113 -0
- data/examples/latex_templates/Test-springer_article/svjour3.cls +1431 -0
- data/examples/misc/moneyball.rb +1 -1
- data/examples/misc/subsetting.rb +37 -37
- data/examples/rmarkdown/svm-rmarkdown-anon-ms-example/svm-rmarkdown-anon-ms-example.Rmd +73 -0
- data/examples/rmarkdown/svm-rmarkdown-anon-ms-example/svm-rmarkdown-anon-ms-example.pdf +0 -0
- data/examples/rmarkdown/svm-rmarkdown-article-example/svm-rmarkdown-article-example.Rmd +382 -0
- data/examples/rmarkdown/svm-rmarkdown-article-example/svm-rmarkdown-article-example.pdf +0 -0
- data/examples/rmarkdown/svm-rmarkdown-beamer-example/svm-rmarkdown-beamer-example.Rmd +164 -0
- data/examples/rmarkdown/svm-rmarkdown-beamer-example/svm-rmarkdown-beamer-example.pdf +0 -0
- data/examples/rmarkdown/svm-rmarkdown-cv/svm-rmarkdown-cv.Rmd +92 -0
- data/examples/rmarkdown/svm-rmarkdown-cv/svm-rmarkdown-cv.pdf +0 -0
- data/examples/rmarkdown/svm-rmarkdown-syllabus-example/attend-grade-relationships.csv +482 -0
- data/examples/rmarkdown/svm-rmarkdown-syllabus-example/svm-rmarkdown-syllabus-example.Rmd +280 -0
- data/examples/rmarkdown/svm-rmarkdown-syllabus-example/svm-rmarkdown-syllabus-example.pdf +0 -0
- data/examples/rmarkdown/svm-xaringan-example/svm-xaringan-example.Rmd +386 -0
- data/lib/R_interface/r.rb +2 -2
- data/lib/R_interface/r_libs.R +6 -1
- data/lib/R_interface/r_methods.rb +12 -2
- data/lib/R_interface/rdata_frame.rb +8 -17
- data/lib/R_interface/rindexed_object.rb +1 -2
- data/lib/R_interface/rlist.rb +1 -0
- data/lib/R_interface/robject.rb +20 -23
- data/lib/R_interface/rpkg.rb +15 -6
- data/lib/R_interface/rsupport.rb +13 -19
- data/lib/R_interface/ruby_extensions.rb +14 -18
- data/lib/R_interface/rvector.rb +0 -12
- data/lib/gknit.rb +2 -0
- data/lib/gknit/draft.rb +105 -0
- data/lib/gknit/knitr_engine.rb +6 -37
- data/lib/util/exec_ruby.rb +22 -84
- data/lib/util/inline_file.rb +7 -3
- data/specs/figures/bg.jpeg +0 -0
- data/specs/figures/bg.png +0 -0
- data/specs/figures/bg.svg +2 -2
- data/specs/figures/dose_len.png +0 -0
- data/specs/figures/no_args.jpeg +0 -0
- data/specs/figures/no_args.png +0 -0
- data/specs/figures/no_args.svg +2 -2
- data/specs/figures/width_height.jpeg +0 -0
- data/specs/figures/width_height.png +0 -0
- data/specs/figures/width_height_units1.jpeg +0 -0
- data/specs/figures/width_height_units1.png +0 -0
- data/specs/figures/width_height_units2.jpeg +0 -0
- data/specs/figures/width_height_units2.png +0 -0
- data/specs/r_dataframe.spec.rb +184 -11
- data/specs/r_list.spec.rb +4 -4
- data/specs/r_list_apply.spec.rb +11 -10
- data/specs/ruby_expression.spec.rb +3 -11
- data/specs/tmp.rb +106 -34
- data/version.rb +1 -1
- metadata +96 -33
- data/bin/gknit_old_r +0 -236
- data/blogs/dev/dev.Rmd +0 -77
- data/blogs/dev/dev.md +0 -87
- data/blogs/dev/dev_files/figure-html/bubble-1.png +0 -0
- data/blogs/dev/dev_files/figure-html/diverging_bar. +0 -0
- data/blogs/dev/dev_files/figure-html/diverging_bar.png +0 -0
- data/blogs/dplyr/dplyr.rb +0 -63
- data/blogs/galaaz_ggplot/galaaz_ggplot.aux +0 -43
- data/blogs/galaaz_ggplot/galaaz_ggplot.log +0 -640
- data/blogs/galaaz_ggplot/galaaz_ggplot.out +0 -10
- data/blogs/galaaz_ggplot/galaaz_ggplot.tex +0 -481
- data/blogs/galaaz_ggplot/midwest.png +0 -0
- data/blogs/galaaz_ggplot/scatter_plot.png +0 -0
- data/blogs/ruby_plot/ruby_plot.Rmd_external_figs +0 -662
- data/blogs/ruby_plot/ruby_plot.tex +0 -1077
- data/blogs/ruby_plot/ruby_plot_files/figure-html/dose_len.svg +0 -57
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facet_by_delivery.svg +0 -106
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facet_by_dose.svg +0 -110
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_by_delivery_color.svg +0 -174
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_by_delivery_color2.svg +0 -236
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_with_jitter.svg +0 -296
- data/blogs/ruby_plot/ruby_plot_files/figure-html/facets_with_points.svg +0 -236
- data/blogs/ruby_plot/ruby_plot_files/figure-html/final_box_plot.svg +0 -218
- data/blogs/ruby_plot/ruby_plot_files/figure-html/final_violin_plot.svg +0 -128
- data/blogs/ruby_plot/ruby_plot_files/figure-html/violin_with_jitter.svg +0 -150
- data/examples/paper/paper.rb +0 -36
Binary file
|
data/blogs/gknit/lst.rds
ADDED
Binary file
|
@@ -0,0 +1,27 @@
|
|
1
|
+
@book{Wilkinson:grammar_of_graphics,
|
2
|
+
author = {Wilkinson, Leland},
|
3
|
+
title = {The Grammar of Graphics (Statistics and Computing)},
|
4
|
+
year = {2005},
|
5
|
+
isbn = {0387245448},
|
6
|
+
publisher = {Springer-Verlag},
|
7
|
+
address = {Berlin, Heidelberg},
|
8
|
+
}
|
9
|
+
|
10
|
+
@article{Knuth:literate_programming,
|
11
|
+
author = {Knuth, Donald E.},
|
12
|
+
title = {Literate Programming},
|
13
|
+
journal = {Comput. J.},
|
14
|
+
issue_date = {May 1984},
|
15
|
+
volume = {27},
|
16
|
+
number = {2},
|
17
|
+
month = may,
|
18
|
+
year = {1984},
|
19
|
+
issn = {0010-4620},
|
20
|
+
pages = {97--111},
|
21
|
+
numpages = {15},
|
22
|
+
url = {http://dx.doi.org/10.1093/comjnl/27.2.97},
|
23
|
+
doi = {10.1093/comjnl/27.2.97},
|
24
|
+
acmid = {479},
|
25
|
+
publisher = {Oxford University Press},
|
26
|
+
address = {Oxford, UK},
|
27
|
+
}
|
Binary file
|
data/blogs/manual/manual.Rmd
CHANGED
@@ -4,27 +4,28 @@ subtitle: "How to tightly couple Ruby and R in GraalVM"
|
|
4
4
|
author: "Rodrigo Botafogo"
|
5
5
|
tags: [Galaaz, Ruby, R, TruffleRuby, FastR, GraalVM, ggplot2]
|
6
6
|
date: "2019"
|
7
|
+
bibliography: "/home/rbotafogo/Bibliography/stats.bib"
|
7
8
|
output:
|
8
|
-
html_document:
|
9
|
-
self_contained: true
|
10
|
-
keep_md: true
|
11
|
-
md_document:
|
12
|
-
variant: markdown_github
|
13
9
|
pdf_document:
|
14
10
|
includes:
|
15
11
|
in_header: "../../sty/galaaz.sty"
|
16
12
|
keep_tex: yes
|
17
13
|
number_sections: yes
|
18
14
|
toc: true
|
19
|
-
toc_depth:
|
15
|
+
toc_depth: 3
|
16
|
+
html_document:
|
17
|
+
self_contained: true
|
18
|
+
keep_md: true
|
19
|
+
md_document:
|
20
|
+
variant: markdown_github
|
20
21
|
fontsize: 11pt
|
21
22
|
---
|
22
23
|
|
23
24
|
```{ruby setup, echo=FALSE}
|
25
|
+
R.options(crayon__enabled: false)
|
24
26
|
R.install_and_loads('kableExtra')
|
25
27
|
```
|
26
28
|
|
27
|
-
|
28
29
|
# Introduction
|
29
30
|
|
30
31
|
Galaaz is a system for tightly coupling Ruby and R. Ruby is a powerful language, with a large
|
@@ -34,6 +35,92 @@ other hand, R is considered one of the most powerful languages for solving all o
|
|
34
35
|
problems. Maybe the strongest competitor to R is Python with libraries such as NumPy,
|
35
36
|
Panda, SciPy, SciKit-Learn and a couple more.
|
36
37
|
|
38
|
+
With Galaaz we do not intend to re-implement any of the scientific libraries in R, we allow
|
39
|
+
for very tight coupling between the two languages to the point that the Ruby developer does
|
40
|
+
not need to know that there is an R engine running.
|
41
|
+
|
42
|
+
According to Wikipedia "Ruby is a dynamic, interpreted, reflective, object-oriented,
|
43
|
+
general-purpose programming language. It was designed and developed in the mid-1990s by Yukihiro
|
44
|
+
"Matz" Matsumoto in Japan." It reached high popularity with the development of Ruby on Rails
|
45
|
+
(RoR) by David Heinemeier Hansson. RoR is a web application framework first released
|
46
|
+
around 2005. It makes extensive use of Ruby's metaprogramming features. With RoR,
|
47
|
+
Ruby became very popular. According to [Ruby's Tiobe index](https://www.tiobe.com/tiobe-index/ruby/)
|
48
|
+
it peeked in popularity around 2008, then declined until 2015 when it started picking up again.
|
49
|
+
At the time of this writing (November 2018), the Tiobe index puts Ruby in 16th position as
|
50
|
+
most popular language.
|
51
|
+
|
52
|
+
Python, a language similar to Ruby, ranks 4th in the index. Java, C and C++ take the
|
53
|
+
first three positions. Ruby is often criticized for its focus on web applications.
|
54
|
+
But Ruby can do [much more](https://github.com/markets/awesome-ruby) than just web applications.
|
55
|
+
Yet, for scientific computing, Ruby lags way behind Python and R. Python has
|
56
|
+
Django framework for web, NumPy for numerical arrays, Pandas for data analysis.
|
57
|
+
R is a free software environment for statistical computing and graphics with thousands
|
58
|
+
of libraries for data analysis.
|
59
|
+
|
60
|
+
Until recently, there was no real perspective for Ruby to bridge this gap.
|
61
|
+
Implementing a complete scientific computing infrastructure would take too long.
|
62
|
+
Enters [Oracle's GraalVM](https://www.graalvm.org/):
|
63
|
+
|
64
|
+
> GraalVM is a universal virtual machine for running applications written in
|
65
|
+
> JavaScript, Python 3, Ruby, R, JVM-based languages like Java, Scala, Kotlin,
|
66
|
+
> and LLVM-based languages such as C and C++.
|
67
|
+
>
|
68
|
+
> GraalVM removes the isolation between programming languages and enables
|
69
|
+
> interoperability in a shared runtime. It can run either standalone or in the
|
70
|
+
> context of OpenJDK, Node.js, Oracle Database, or MySQL.
|
71
|
+
>
|
72
|
+
> GraalVM allows you to write polyglot applications with a seamless way to pass
|
73
|
+
> values from one language to another. With GraalVM there is no copying or
|
74
|
+
> marshaling necessary as it is with other polyglot systems. This lets you
|
75
|
+
> achieve high performance when language boundaries are crossed. Most of the time
|
76
|
+
> there is no additional cost for crossing a language boundary at all.
|
77
|
+
>
|
78
|
+
> Often developers have to make uncomfortable compromises that require them
|
79
|
+
> to rewrite their software in other languages. For example:
|
80
|
+
>
|
81
|
+
> * That library is not available in my language. I need to rewrite it.
|
82
|
+
> * That language would be the perfect fit for my problem, but we cannot
|
83
|
+
> run it in our environment.
|
84
|
+
> * That problem is already solved in my language, but the language is
|
85
|
+
> too slow.
|
86
|
+
>
|
87
|
+
> With GraalVM we aim to allow developers to freely choose the right language for
|
88
|
+
> the task at hand without making compromises.
|
89
|
+
|
90
|
+
As stated above, GraalVM is a _universal_ virtual machine that allows Ruby and R (and other
|
91
|
+
languages) to run on the same environment. GraalVM allows polyglot applications to
|
92
|
+
_seamlessly_ interact with one another and pass values from one language to the other.
|
93
|
+
Although a great idea, GraalVM still requires application writers to know several languages.
|
94
|
+
To eliminate that requirement, we built Galaaz, a gem for Ruby, to tightly couple
|
95
|
+
Ruby and R and allow those languages to interact in a way that the user will be unaware
|
96
|
+
of such interaction. In other words, a Ruby programmer will be able to use all
|
97
|
+
the capabilities of R without knowing the R syntax.
|
98
|
+
|
99
|
+
Library wrapping is a usual way of bringing features from one language into another.
|
100
|
+
To improve performance, Python often wraps more efficient C libraries. For the
|
101
|
+
Python developer, the existence of such C libraries is hidden. The problem with
|
102
|
+
library wrapping is that for any new library, there is the need to handcraft a new
|
103
|
+
wrapper.
|
104
|
+
|
105
|
+
Galaaz, instead of wrapping a single C or R library, wraps the whole R language
|
106
|
+
in Ruby. Doing so, all thousands of R libraries are available immediately
|
107
|
+
to Ruby developers without any new wrapping effort.
|
108
|
+
|
109
|
+
## What does Galaaz mean
|
110
|
+
|
111
|
+
Galaaz is the Portuguese name for "Galahad". From Wikipedia:
|
112
|
+
|
113
|
+
Sir Galahad (sometimes referred to as Galeas or Galath),
|
114
|
+
in Arthurian legend, is a knight of King Arthur's Round Table and one
|
115
|
+
of the three achievers of the Holy Grail. He is the illegitimate son
|
116
|
+
of Sir Lancelot and Elaine of Corbenic, and is renowned for his
|
117
|
+
gallantry and purity as the most perfect of all knights. Emerging quite
|
118
|
+
late in the medieval Arthurian tradition, Sir Galahad first appears in the
|
119
|
+
Lancelot–Grail cycle, and his story is taken up in later works such as
|
120
|
+
the Post-Vulgate Cycle and Sir Thomas Malory's Le Morte d'Arthur.
|
121
|
+
His name should not be mistaken with Galehaut, a different knight from
|
122
|
+
Arthurian legend.
|
123
|
+
|
37
124
|
# System Compatibility
|
38
125
|
|
39
126
|
* Oracle Linux 7
|
@@ -84,7 +171,7 @@ Panda, SciPy, SciKit-Learn and a couple more.
|
|
84
171
|
> galaaz -T
|
85
172
|
|
86
173
|
Shows a list with all available executalbe tasks. To execute a task, substitute the
|
87
|
-
|
174
|
+
'rake' word in the list with 'galaaz'. For instance, the following line shows up
|
88
175
|
after 'galaaz -T'
|
89
176
|
|
90
177
|
rake master_list:scatter_plot # scatter_plot from:....
|
@@ -93,9 +180,711 @@ Panda, SciPy, SciKit-Learn and a couple more.
|
|
93
180
|
|
94
181
|
> galaaz master_list:scatter_plot
|
95
182
|
|
96
|
-
# Basic Types
|
97
183
|
|
98
|
-
|
184
|
+
# Accessing R from Ruby
|
185
|
+
|
186
|
+
One of the nice aspects of Galaaz on GraalVM, is that variables and functions defined in R, can
|
187
|
+
be easily accessed from Ruby. For instance, to access the 'mtcars' data frame from R
|
188
|
+
in Ruby, we use the ':mtcar' symbol preceded by the '~' operator, thus '~:r_vec' retrieves the
|
189
|
+
value of the 'mtcars' variable.
|
190
|
+
|
191
|
+
```{ruby access_r}
|
192
|
+
puts ~:mtcars
|
193
|
+
```
|
194
|
+
|
195
|
+
To access an R function from Ruby, the R function needs to be preceeded by 'R.' scoping.
|
196
|
+
Bellow we see and example of creating a R::Vector by calling the 'c' R function
|
197
|
+
|
198
|
+
```{ruby call_r_func}
|
199
|
+
puts vec = R.c(1.0, 2.0, 3.0, 4.0)
|
200
|
+
```
|
201
|
+
Note that 'vec' is an object of type R::Vector:
|
202
|
+
|
203
|
+
```{ruby r_object}
|
204
|
+
puts vec.class
|
205
|
+
```
|
206
|
+
Every object created by a call to an R function will be of a type that inherits from
|
207
|
+
R::Object. In R, there is also a function 'class'. In order to access that function we
|
208
|
+
can call method 'rclass' in the R::Object:
|
209
|
+
|
210
|
+
```{ruby rclass}
|
211
|
+
puts vec.rclass
|
212
|
+
```
|
213
|
+
When working with R::Object(s), it is possible to use the '.' operator to pipe operations.
|
214
|
+
When using '.', the object to which the '.' is applied becomes the first argument of the
|
215
|
+
corresponding R function. For instance, function 'c' in R, can be used to concatenate
|
216
|
+
two vectors or more vectors (in R, there are no scalar values, scalars are converted to
|
217
|
+
vectors of size 1. Within Galaaz, scalar parameter is converted to a size one vector):
|
218
|
+
|
219
|
+
```{ruby concat}
|
220
|
+
puts R.c(vec, 10, 20, 30)
|
221
|
+
```
|
222
|
+
The call above to the 'c' function can also be done using '.' notation:
|
223
|
+
|
224
|
+
```{ruby concat_with_dot}
|
225
|
+
puts vec.c(10, 20, 30)
|
226
|
+
```
|
227
|
+
We will talk about vector indexing in a latter section. But notice here that indexing
|
228
|
+
an R::Vector will return another R::Vector:
|
229
|
+
|
230
|
+
```{ruby indexing}
|
231
|
+
puts vec[1]
|
232
|
+
```
|
233
|
+
Sometimes we want to index an R::Object and get back a Ruby object that is not wrapped
|
234
|
+
in an R::Object, but the native Ruby object. For this, we can index the R object with
|
235
|
+
the '>>' operator:
|
236
|
+
|
237
|
+
```{ruby native_value}
|
238
|
+
puts vec >> 0
|
239
|
+
puts vec >> 2
|
240
|
+
```
|
241
|
+
|
242
|
+
It is also possible to call an R function with named arguments, by creating the function
|
243
|
+
in Galaaz with named parameters. For instance, here is an example of creating a 'list'
|
244
|
+
with named elements:
|
245
|
+
|
246
|
+
```{ruby named_parameters}
|
247
|
+
puts R.list(first_name: "Rodrigo", last_name: "Botafogo")
|
248
|
+
```
|
249
|
+
|
250
|
+
Many R functions receive another function as argument. For instance, method 'map' applies
|
251
|
+
a function to every element of a vector. With Galaaz, it is possible to pass a Proc,
|
252
|
+
Method or Lambda in place of the expected R function. In this next example, we will
|
253
|
+
add 2 to every element of our previously created vector:
|
254
|
+
|
255
|
+
```{ruby proc_as_param}
|
256
|
+
puts vec.map { |x| x + 2 }
|
257
|
+
```
|
258
|
+
|
259
|
+
# gKnitting a Document
|
260
|
+
|
261
|
+
This manual has been formatted usign gKnit. gKnit uses Knitr and R markdown to knit
|
262
|
+
a document in Ruby or R and output it in any of the available formats for R markdown.
|
263
|
+
gKnit runs atop of GraalVM, and Galaaz. In gKnit, Ruby variables are persisted between
|
264
|
+
chunks, making it an ideal solution for literate programming. Also, since it is based
|
265
|
+
on Galaaz, Ruby chunks can have access to R variables and Polyglot Programming with
|
266
|
+
Ruby and R is quite natural.
|
267
|
+
|
268
|
+
The idea of "literate programming" was first introduced by Donald Knuth in the
|
269
|
+
1980's [@Knuth:literate_programming].
|
270
|
+
The main intention of this approach was to develop software interspersing macro snippets,
|
271
|
+
traditional source code, and a natural language such as English in a document
|
272
|
+
that could be compiled into
|
273
|
+
executable code and at the same time easily read by a human developer. According to Knuth
|
274
|
+
"The practitioner of
|
275
|
+
literate programming can be regarded as an essayist, whose main concern is with exposition
|
276
|
+
and excellence of style."
|
277
|
+
|
278
|
+
The idea of literate programming evolved into the idea of reproducible research, in which
|
279
|
+
all the data, software code, documentation, graphics etc. needed to reproduce the research
|
280
|
+
and its reports could be included in a
|
281
|
+
single document or set of documents that when distributed to peers could be rerun generating
|
282
|
+
the same output and reports.
|
283
|
+
|
284
|
+
The R community has put a great deal of effort in reproducible research. In 2002, Sweave was
|
285
|
+
introduced and it allowed mixing R code with Latex generating high quality PDF documents. A
|
286
|
+
Sweave document could include code, the results of executing the code, graphics and text
|
287
|
+
such that it contained the whole narrative to reproduce the research. In
|
288
|
+
2012, Knitr, developed by Yihui Xie from RStudio was released to replace Sweave and to
|
289
|
+
consolidate in one single package the many extensions and add-on packages that
|
290
|
+
were necessary for Sweave.
|
291
|
+
|
292
|
+
With Knitr, __R markdown__ was also developed, an extension to the
|
293
|
+
Markdown format. With __R markdown__ and Knitr it is possible to generate reports in a multitude
|
294
|
+
of formats such as HTML, markdown, Latex, PDF, dvi, etc. __R markdown__ also allows the use of
|
295
|
+
multiple programming languages such as R, Ruby, Python, etc. in the same document.
|
296
|
+
|
297
|
+
In __R markdown__, text is interspersed with
|
298
|
+
code chunks that can be executed and both the code and its results can become
|
299
|
+
part of the final report. Although __R markdown__ allows multiple programming languages in the
|
300
|
+
same document, only R and Python (with
|
301
|
+
the reticulate package) can persist variables between chunks. For other languages, such as
|
302
|
+
Ruby, every chunk will start a new process and thus all data is lost between chunks, unless it
|
303
|
+
is somehow stored in a data file that is read by the next chunk.
|
304
|
+
|
305
|
+
Being able to persist data
|
306
|
+
between chunks is critical for literate programming otherwise the flow of the narrative is lost
|
307
|
+
by all the effort of having to save data and then reload it. Although this might, at first, seem like
|
308
|
+
a small nuisance, not being able to persist data between chunks is a major issue. For example, let's
|
309
|
+
take a look at the following simple example in which we want to show how to create a list and the
|
310
|
+
use it. Let's first assume that data cannot be persisted between chunks. In the next chunk we
|
311
|
+
create a list, then we would need to save it to file, but to save it, we need somehow to marshal the
|
312
|
+
data into a binary format:
|
313
|
+
|
314
|
+
```{ruby no_persistence}
|
315
|
+
lst = R.list(a: 1, b: 2, c: 3)
|
316
|
+
lst.saveRDS("lst.rds")
|
317
|
+
```
|
318
|
+
then, on the next chunk, where variable 'lst' is used, we need to read back it's value
|
319
|
+
|
320
|
+
```{ruby load_persisted_data}
|
321
|
+
lst = R.readRDS("lst.rds")
|
322
|
+
puts lst
|
323
|
+
```
|
324
|
+
|
325
|
+
Now, any single code has dozens of variables that we might want to use and reuse between chunks.
|
326
|
+
Clearly, such an approach becomes quickly unmanageable. Probably, because of
|
327
|
+
this problem, it is very rare to see any __R markdown__ document in the Ruby community.
|
328
|
+
|
329
|
+
When variables can be used accross chunks, then no overhead is needed:
|
330
|
+
|
331
|
+
```{ruby persistence}
|
332
|
+
lst = R.list(a: 1, b: 2, c: 3)
|
333
|
+
# any other code can be added here
|
334
|
+
```
|
335
|
+
|
336
|
+
```{ruby use_var}
|
337
|
+
puts lst
|
338
|
+
```
|
339
|
+
|
340
|
+
In the Python community, the same effort to have code and text in an integrated environment
|
341
|
+
started around the first decade of 2000. In 2006 iPython 0.7.2 was released. In 2014,
|
342
|
+
Fernando Pérez, spun off project Jupyter from iPython creating a web-based interactive
|
343
|
+
computation environment. Jupyter can now be used with many languages, including Ruby with the
|
344
|
+
iruby gem (https://github.com/SciRuby/iruby). In order to have multiple languages in a Jupyter
|
345
|
+
notebook the SoS kernel was developed (https://vatlab.github.io/sos-docs/).
|
346
|
+
|
347
|
+
## gKnit and __R markdown__
|
348
|
+
|
349
|
+
gKnit is based on knitr and __R markdown__ and can knit a document
|
350
|
+
written both in Ruby and/or R and output it in any of the available formats of __R markdown__. gKnit
|
351
|
+
allows ruby developers to do literate programming and reproducible research by allowing them to
|
352
|
+
have in a single document, text and code.
|
353
|
+
|
354
|
+
In gKnit, Ruby variables are persisted between
|
355
|
+
chunks, making it an ideal solution for literate programming in this language. Also,
|
356
|
+
since it is based on Galaaz, Ruby chunks can have access to R variables and Polyglot Programming
|
357
|
+
with Ruby and R is quite natural.
|
358
|
+
|
359
|
+
This is not a blog post on __R markdown__, and the interested user is directed to the following links
|
360
|
+
for detailed information on its capabilities and use.
|
361
|
+
|
362
|
+
* https://rmarkdown.rstudio.com/ or
|
363
|
+
* https://bookdown.org/yihui/rmarkdown/
|
364
|
+
|
365
|
+
In this post, we will describe just the main aspects of __R markdown__, so the user can start
|
366
|
+
gKnitting Ruby and R documents quickly.
|
367
|
+
|
368
|
+
## The Yaml header
|
369
|
+
|
370
|
+
An __R markdown__ document should start with a Yaml header and be stored in a file with
|
371
|
+
'.Rmd' extension. This document has the following header for gKitting an HTML document.
|
372
|
+
|
373
|
+
```
|
374
|
+
---
|
375
|
+
title: "How to do reproducible research in Ruby with gKnit"
|
376
|
+
author:
|
377
|
+
- "Rodrigo Botafogo"
|
378
|
+
- "Daniel Mossé - University of Pittsburgh"
|
379
|
+
tags: [Tech, Data Science, Ruby, R, GraalVM]
|
380
|
+
date: "20/02/2019"
|
381
|
+
output:
|
382
|
+
html_document:
|
383
|
+
self_contained: true
|
384
|
+
keep_md: true
|
385
|
+
pdf_document:
|
386
|
+
includes:
|
387
|
+
in_header: ["../../sty/galaaz.sty"]
|
388
|
+
number_sections: yes
|
389
|
+
---
|
390
|
+
```
|
391
|
+
|
392
|
+
For more information on the options in the Yaml header, [check here](https://bookdown.org/yihui/rmarkdown/html-document.html).
|
393
|
+
|
394
|
+
## __R Markdown__ formatting
|
395
|
+
|
396
|
+
Document formatting can be done with simple markups such as:
|
397
|
+
|
398
|
+
## Headers
|
399
|
+
|
400
|
+
```
|
401
|
+
# Header 1
|
402
|
+
|
403
|
+
## Header 2
|
404
|
+
|
405
|
+
### Header 3
|
406
|
+
|
407
|
+
```
|
408
|
+
|
409
|
+
## Lists
|
410
|
+
|
411
|
+
```
|
412
|
+
Unordered lists:
|
413
|
+
|
414
|
+
* Item 1
|
415
|
+
* Item 2
|
416
|
+
+ Item 2a
|
417
|
+
+ Item 2b
|
418
|
+
```
|
419
|
+
|
420
|
+
```
|
421
|
+
Ordered Lists
|
422
|
+
|
423
|
+
1. Item 1
|
424
|
+
2. Item 2
|
425
|
+
3. Item 3
|
426
|
+
+ Item 3a
|
427
|
+
+ Item 3b
|
428
|
+
```
|
429
|
+
|
430
|
+
For more R markdown formatting go to https://rmarkdown.rstudio.com/authoring_basics.html.
|
431
|
+
|
432
|
+
## R chunks
|
433
|
+
|
434
|
+
Running and executing Ruby and R code is actually what really interests us is this blog.
|
435
|
+
Inserting a code chunk is done by adding code in a block delimited by three back ticks
|
436
|
+
followed by an open
|
437
|
+
curly brace ('{') followed with the engine name (r, ruby, rb, include, ...), an
|
438
|
+
any optional chunk_label and options, as shown bellow:
|
439
|
+
|
440
|
+
````
|
441
|
+
```{engine_name [chunk_label], [chunk_options]}`r ''`
|
442
|
+
```
|
443
|
+
````
|
444
|
+
|
445
|
+
for instance, let's add an R chunk to the document labeled 'first_r_chunk'. This is
|
446
|
+
a very simple code just to create a variable and print it out, as follows:
|
447
|
+
|
448
|
+
````
|
449
|
+
```{r first_r_chunk}`r ''`
|
450
|
+
vec <- c(1, 2, 3)
|
451
|
+
print(vec)
|
452
|
+
```
|
453
|
+
````
|
454
|
+
|
455
|
+
If this block is added to an __R markdown__ document and gKnitted the result will be:
|
456
|
+
|
457
|
+
```{r first_r_chunk}
|
458
|
+
vec <- c(1, 2, 3)
|
459
|
+
print(vec)
|
460
|
+
```
|
461
|
+
|
462
|
+
Now let's say that we want to do some analysis in the code, but just print the result and not the
|
463
|
+
code itself. For this, we need to add the option 'echo = FALSE'.
|
464
|
+
|
465
|
+
````
|
466
|
+
```{r second_r_chunk, echo = FALSE}`r ''`
|
467
|
+
vec2 <- c(10, 20, 30)
|
468
|
+
vec3 <- vec * vec2
|
469
|
+
print(vec3)
|
470
|
+
```
|
471
|
+
````
|
472
|
+
Here is how this block will show up in the document. Observe that the code is not shown
|
473
|
+
and we only see the execution result in a white box
|
474
|
+
|
475
|
+
```{r second_r_chunk, echo = FALSE}
|
476
|
+
vec2 <- c(10, 20, 30)
|
477
|
+
vec3 <- vec * vec2
|
478
|
+
print(vec3)
|
479
|
+
```
|
480
|
+
|
481
|
+
A description of the available chunk options can be found in https://yihui.name/knitr/.
|
482
|
+
|
483
|
+
Let's add another R chunk with a function definition. In this example, a vector
|
484
|
+
'r_vec' is created and
|
485
|
+
a new function 'reduce_sum' is defined. The chunk specification is
|
486
|
+
|
487
|
+
````
|
488
|
+
```{r data_creation}`r ''`
|
489
|
+
r_vec <- c(1, 2, 3, 4, 5)
|
490
|
+
|
491
|
+
reduce_sum <- function(...) {
|
492
|
+
Reduce(sum, as.list(...))
|
493
|
+
}
|
494
|
+
```
|
495
|
+
````
|
496
|
+
|
497
|
+
and this is how it will look like once executed. From now on, to be concise in the
|
498
|
+
presentation we will not show chunk definitions any longer.
|
499
|
+
|
500
|
+
|
501
|
+
```{r data_creation}
|
502
|
+
r_vec <- c(1, 2, 3, 4, 5)
|
503
|
+
|
504
|
+
reduce_sum <- function(...) {
|
505
|
+
Reduce(sum, as.list(...))
|
506
|
+
}
|
507
|
+
```
|
508
|
+
|
509
|
+
We can, possibly in another chunk, access the vector and call the function as follows:
|
510
|
+
|
511
|
+
```{r using_previous}
|
512
|
+
print(r_vec)
|
513
|
+
print(reduce_sum(r_vec))
|
514
|
+
```
|
515
|
+
## R Graphics with ggplot
|
516
|
+
|
517
|
+
In the following chunk, we create a bubble chart in R using ggplot and include it in
|
518
|
+
this document. Note that there is no directive in the code to include the image, this
|
519
|
+
occurs automatically. The 'mpg' dataframe is natively available to R and to Galaaz as
|
520
|
+
well.
|
521
|
+
|
522
|
+
For the reader not knowledgeable of ggplot, ggplot is a graphics library based on "the
|
523
|
+
grammar of graphics" [@Wilkinson:grammar_of_graphics]. The idea of the grammar of graphics
|
524
|
+
is to build a graphics by adding layers to the plot. More information can be found in
|
525
|
+
https://towardsdatascience.com/a-comprehensive-guide-to-the-grammar-of-graphics-for-effective-visualization-of-multi-dimensional-1f92b4ed4149.
|
526
|
+
|
527
|
+
In the plot bellow the 'mpg' dataset from base R is used. "The data concerns city-cycle fuel
|
528
|
+
consumption in miles per gallon, to be predicted in terms of 3 multivalued discrete and 5
|
529
|
+
continuous attributes." (Quinlan, 1993)
|
530
|
+
|
531
|
+
First, the 'mpg' dataset if filtered to extract only cars from the following manumactures: Audi, Ford,
|
532
|
+
Honda, and Hyundai and stored in the 'mpg_select' variable. Then, the selected dataframe is passed
|
533
|
+
to the ggplot function specifying in the aesthetic method (aes) that 'displacement' (disp) should
|
534
|
+
be plotted in the 'x' axis and 'city mileage' should be on the 'y' axis. In the 'labs' layer we
|
535
|
+
pass the 'title' and 'subtitle' for the plot. To the basic plot 'g', geom\_jitter is added, that
|
536
|
+
plots cars from the same manufactures with the same color (col=manufactures) and the size of the
|
537
|
+
car point equal its high way consumption (size = hwy). Finally, a last layer is plotter containing
|
538
|
+
a linear regression line (method = "lm") for every manufacturer.
|
539
|
+
|
540
|
+
```{r bubble, dev='png'}
|
541
|
+
# load package and data
|
542
|
+
library(ggplot2)
|
543
|
+
data(mpg, package="ggplot2")
|
544
|
+
|
545
|
+
mpg_select <- mpg[mpg$manufacturer %in% c("audi", "ford", "honda", "hyundai"), ]
|
546
|
+
|
547
|
+
# Scatterplot
|
548
|
+
theme_set(theme_bw()) # pre-set the bw theme.
|
549
|
+
g <- ggplot(mpg_select, aes(displ, cty)) +
|
550
|
+
labs(subtitle="mpg: Displacement vs City Mileage",
|
551
|
+
title="Bubble chart")
|
552
|
+
|
553
|
+
g + geom_jitter(aes(col=manufacturer, size=hwy)) +
|
554
|
+
geom_smooth(aes(col=manufacturer), method="lm", se=F)
|
555
|
+
```
|
556
|
+
|
557
|
+
## Ruby chunks
|
558
|
+
|
559
|
+
Including a Ruby chunk is just as easy as including an R chunk in the document: just
|
560
|
+
change the name of the engine to 'ruby'. It is also possible to pass chunk options
|
561
|
+
to the Ruby engine; however, this version does not accept all the options that are
|
562
|
+
available to R chunks. Future versions will add those options.
|
563
|
+
|
564
|
+
````
|
565
|
+
```{ruby first_ruby_chunk}`r ''`
|
566
|
+
```
|
567
|
+
````
|
568
|
+
|
569
|
+
In this example, the ruby chunk is called 'first_ruby_chunk'. One important
|
570
|
+
aspect of chunk labels is that they cannot be duplicated. If a chunk label is
|
571
|
+
duplicated, gKnit will stop with an error.
|
572
|
+
|
573
|
+
In the following chunk, variable 'a', 'b' and 'c' are standard Ruby variables
|
574
|
+
and 'vec' and 'vec2' are two vectors created by calling the 'c' method on the
|
575
|
+
R module.
|
576
|
+
|
577
|
+
In Galaaz, the R module allows us to access R functions transparently. The 'c'
|
578
|
+
function in R, is a function that concatenates its arguments making a vector.
|
579
|
+
|
580
|
+
It
|
581
|
+
should be clear that there is no requirement in gknit to call or use any R
|
582
|
+
functions. gKnit will knit standard Ruby code, or even general text without
|
583
|
+
any code.
|
584
|
+
|
585
|
+
```{ruby split_data}
|
586
|
+
a = [1, 2, 3]
|
587
|
+
b = "US$ 250.000"
|
588
|
+
c = "The 'outputs' function"
|
589
|
+
|
590
|
+
vec = R.c(1, 2, 3)
|
591
|
+
vec2 = R.c(10, 20, 30)
|
592
|
+
```
|
593
|
+
|
594
|
+
In the next block, variables 'a', 'vec' and 'vec2' are used and printed.
|
595
|
+
|
596
|
+
```{ruby split2}
|
597
|
+
puts a
|
598
|
+
puts vec * vec2
|
599
|
+
```
|
600
|
+
|
601
|
+
Note that 'a' is a standard Ruby Array and 'vec' and 'vec2' are vectors that behave accordingly,
|
602
|
+
where multiplication works as expected.
|
603
|
+
|
604
|
+
## Inline Ruby code
|
605
|
+
|
606
|
+
When using a Ruby chunk, the code and the output are formatted in blocks as seen above.
|
607
|
+
This formatting is not always desired. Sometimes, we want to have the results of the
|
608
|
+
Ruby evaluation included in the middle of a phrase. gKnit allows adding inline Ruby code
|
609
|
+
with the 'rb' engine. The following chunk specification will
|
610
|
+
create and inline Ruby text:
|
611
|
+
|
612
|
+
````
|
613
|
+
This is some text with inline Ruby accessing variable 'b' which has value:
|
614
|
+
```{rb puts "```{rb puts b}\n```"}
|
615
|
+
```
|
616
|
+
and is followed by some other text!
|
617
|
+
````
|
618
|
+
|
619
|
+
<div style="margin-bottom:30px;">
|
620
|
+
</div>
|
621
|
+
|
622
|
+
This is some text with inline Ruby accessing variable 'b' which has value:
|
623
|
+
```{rb puts b}
|
624
|
+
```
|
625
|
+
and is followed by some other text!
|
626
|
+
|
627
|
+
<div style="margin-bottom:30px;">
|
628
|
+
</div>
|
629
|
+
|
630
|
+
Note that it is important not to add any new line before of after the code
|
631
|
+
block if we want everything to be in only one line, resulting in the following sentence
|
632
|
+
with inline Ruby code.
|
633
|
+
|
634
|
+
|
635
|
+
```{ruby heading, echo = FALSE}
|
636
|
+
outputs "### #{c}"
|
637
|
+
```
|
638
|
+
|
639
|
+
He have previously used the standard 'puts' method in Ruby chunks in order produce
|
640
|
+
output. The result of a 'puts', as seen in all previous chunks that use it, is formatted
|
641
|
+
inside a white box that
|
642
|
+
follows the code block. Many times however, we would like to do some processing in the
|
643
|
+
Ruby chunk and have the result of this processing generate and output that is
|
644
|
+
"included" in the document as if we had typed it in __R markdown__ document.
|
645
|
+
|
646
|
+
For example, suppose we want to create a new heading in our document, but the heading
|
647
|
+
phrase is the result of some code processing: maybe it's the first line of a file we are
|
648
|
+
going to read. Method 'outputs' adds its output as if typed in the __R markdown__ document.
|
649
|
+
|
650
|
+
Take now a look at variable 'c' (it was defined in a previous block above) as
|
651
|
+
'c = "The 'outputs' function". "The 'outputs' function" is actually the name of this
|
652
|
+
section and it was created using the 'outputs' function inside a Ruby chunk.
|
653
|
+
|
654
|
+
The ruby chunk to generate this heading is:
|
655
|
+
|
656
|
+
````
|
657
|
+
```{ruby heading}`r ''`
|
658
|
+
outputs "### #{c}"
|
659
|
+
```
|
660
|
+
````
|
661
|
+
|
662
|
+
The three '###' is the way we add a Heading 3 in __R markdown__.
|
663
|
+
|
664
|
+
|
665
|
+
### HTML Output from Ruby Chunks
|
666
|
+
|
667
|
+
We've just seen the use of method 'outputs' to add text to the the __R markdown__
|
668
|
+
document. This technique can also be used to add HTML code to the document. In
|
669
|
+
__R markdown__, any html code typed directly in the document will be properly rendered.
|
670
|
+
Here, for instance, is a table definition in HTML and its output in the document:
|
671
|
+
|
672
|
+
```
|
673
|
+
<table style="width:100%">
|
674
|
+
<tr>
|
675
|
+
<th>Firstname</th>
|
676
|
+
<th>Lastname</th>
|
677
|
+
<th>Age</th>
|
678
|
+
</tr>
|
679
|
+
<tr>
|
680
|
+
<td>Jill</td>
|
681
|
+
<td>Smith</td>
|
682
|
+
<td>50</td>
|
683
|
+
</tr>
|
684
|
+
<tr>
|
685
|
+
<td>Eve</td>
|
686
|
+
<td>Jackson</td>
|
687
|
+
<td>94</td>
|
688
|
+
</tr>
|
689
|
+
</table>
|
690
|
+
```
|
691
|
+
<div style="margin-bottom:30px;">
|
692
|
+
</div>
|
693
|
+
|
694
|
+
<table style="width:100%">
|
695
|
+
<tr>
|
696
|
+
<th>Firstname</th>
|
697
|
+
<th>Lastname</th>
|
698
|
+
<th>Age</th>
|
699
|
+
</tr>
|
700
|
+
<tr>
|
701
|
+
<td>Jill</td>
|
702
|
+
<td>Smith</td>
|
703
|
+
<td>50</td>
|
704
|
+
</tr>
|
705
|
+
<tr>
|
706
|
+
<td>Eve</td>
|
707
|
+
<td>Jackson</td>
|
708
|
+
<td>94</td>
|
709
|
+
</tr>
|
710
|
+
</table>
|
711
|
+
|
712
|
+
<div style="margin-bottom:30px;">
|
713
|
+
</div>
|
714
|
+
|
715
|
+
But manually creating HTML output is not always easy or desirable, specially
|
716
|
+
if we intend the document to be rendered in other formats, for example, as Latex.
|
717
|
+
Also, The above
|
718
|
+
table looks ugly. The 'kableExtra' library is a great library for
|
719
|
+
creating beautiful tables. Take a look at https://cran.r-project.org/web/packages/kableExtra/vignettes/awesome_table_in_html.html
|
720
|
+
|
721
|
+
In the next chunk, we output the 'mtcars' dataframe from R in a nicely formatted
|
722
|
+
table. Note that we retrieve the mtcars dataframe by using '~:mtcars'.
|
723
|
+
|
724
|
+
```{ruby nice_table}
|
725
|
+
R.install_and_loads('kableExtra')
|
726
|
+
outputs (~:mtcars).kable.kable_styling
|
727
|
+
```
|
728
|
+
|
729
|
+
## Including Ruby files in a chunk
|
730
|
+
|
731
|
+
R is a language that was created to be easy and fast for statisticians to use. As far
|
732
|
+
as I know, it was not a
|
733
|
+
language to be used for developing large systems. Of course, there are large systems and
|
734
|
+
libraries in R, but the focus of the language is for developing statistical models and
|
735
|
+
distribute that to peers.
|
736
|
+
|
737
|
+
Ruby on the other hand, is a language for large software development. Systems written in
|
738
|
+
Ruby will have dozens, hundreds or even thousands of files. To document a
|
739
|
+
large system with literate programming, we cannot expect the developer to add all the
|
740
|
+
files in a single '.Rmd' file. gKnit provides the 'include' chunk engine to include
|
741
|
+
a Ruby file as if it had being typed in the '.Rmd' file.
|
742
|
+
|
743
|
+
To include a file, the following chunk should be created, where <filename> is the name of
|
744
|
+
the file to be included and where the extension, if it is '.rb', does not need to be added.
|
745
|
+
If the 'relative' option is not included, then it is treated as TRUE. When 'relative' is
|
746
|
+
true, ruby's 'require\_relative' semantics is used to load the file, when false, Ruby's
|
747
|
+
\$LOAD_PATH is searched to find the file and it is 'require'd.
|
748
|
+
|
749
|
+
````
|
750
|
+
```{include <filename>, relative = <TRUE/FALSE>}`r ''`
|
751
|
+
```
|
752
|
+
````
|
753
|
+
|
754
|
+
Bellow we include file 'model.rb', which is in the same directory of this blog.
|
755
|
+
This code uses R 'caret' package to split a dataset in a train and test sets.
|
756
|
+
The 'caret' package is a very important a useful package for doing Data Analysis,
|
757
|
+
it has hundreds of functions for all steps of the Data Analysis workflow. To
|
758
|
+
use 'caret' just to split a dataset is like using the proverbial cannon to
|
759
|
+
kill the fly. We use it here only to show that integrating Ruby and R and
|
760
|
+
using even a very complex package as 'caret' is trivial with Galaaz.
|
761
|
+
|
762
|
+
A word of advice: the 'caret' package has lots of dependencies and installing
|
763
|
+
it in a Linux system is a time consuming operation. Method 'R.install_and_loads'
|
764
|
+
will install the package if it is not already installed and can take a while.
|
765
|
+
|
766
|
+
````
|
767
|
+
```{include model}`r ''`
|
768
|
+
```
|
769
|
+
````
|
770
|
+
|
771
|
+
```{include model}
|
772
|
+
```
|
773
|
+
|
774
|
+
```{ruby model_partition}
|
775
|
+
mtcars = ~:mtcars
|
776
|
+
model = Model.new(mtcars, percent_train: 0.8)
|
777
|
+
model.partition(:mpg)
|
778
|
+
puts model.train.head
|
779
|
+
puts model.test.head
|
780
|
+
```
|
781
|
+
|
782
|
+
## Documenting Gems
|
783
|
+
|
784
|
+
gKnit also allows developers to document and load files that are not in the same directory
|
785
|
+
of the '.Rmd' file.
|
786
|
+
|
787
|
+
Here is an example of loading the 'find.rb' file from TruffleRuby. In this example, relative
|
788
|
+
is set to FALSE, so Ruby will look for the file in its $LOAD\_PATH, and the user does not
|
789
|
+
need to no it's directory.
|
790
|
+
|
791
|
+
````
|
792
|
+
```{include find, relative = FALSE}`r ''`
|
793
|
+
```
|
794
|
+
````
|
795
|
+
|
796
|
+
```{include find, relative = FALSE}
|
797
|
+
```
|
798
|
+
|
799
|
+
## Converting to PDF
|
800
|
+
|
801
|
+
One of the beauties of knitr is that the same input can be converted to many different outputs.
|
802
|
+
One very useful format, is, of course, PDF. In order to converted an __R markdown__ file to PDF
|
803
|
+
it is necessary to have LaTeX installed on the system. We will not explain here how to
|
804
|
+
install LaTeX as there are plenty of documents on the web showing how to proceed.
|
805
|
+
|
806
|
+
gKnit comes with a simple LaTeX style file for gknitting this blog as a PDF document. Here is
|
807
|
+
the Yaml header to generate this blog in PDF format instead of HTML:
|
808
|
+
|
809
|
+
```
|
810
|
+
---
|
811
|
+
title: "gKnit - Ruby and R Knitting with Galaaz in GraalVM"
|
812
|
+
author: "Rodrigo Botafogo"
|
813
|
+
tags: [Galaaz, Ruby, R, TruffleRuby, FastR, GraalVM, knitr, gknit]
|
814
|
+
date: "29 October 2018"
|
815
|
+
output:
|
816
|
+
pdf\_document:
|
817
|
+
includes:
|
818
|
+
in\_header: ["../../sty/galaaz.sty"]
|
819
|
+
number\_sections: yes
|
820
|
+
---
|
821
|
+
```
|
822
|
+
|
823
|
+
## Template based documents generation
|
824
|
+
|
825
|
+
When a document is converted to PDF it follows a certain convertion template. We've seen above
|
826
|
+
the use of 'galaaz.sty' as a basic template to generate a PDF document. Using the
|
827
|
+
'gknit-draft' app that comes with Galaaz, the same .Rmd file can be compiled to different
|
828
|
+
looking PDF documents. Galaaz automatically loads the 'rticles' R package that comes with
|
829
|
+
templates for the following journals with the respective template name:
|
830
|
+
|
831
|
+
* ACM articles: acm_article
|
832
|
+
* ACS articles: acs_article
|
833
|
+
* AEA journal submissions: aea_article
|
834
|
+
* AGU journal submissions: ????
|
835
|
+
* AMS articles: ams_article
|
836
|
+
* American Statistical Association: asa_article
|
837
|
+
* Biometrics articles: biometrics_article
|
838
|
+
* Bulletin de l'AMQ journal submissions: amq_article
|
839
|
+
* CTeX documents: ctex
|
840
|
+
* Elsevier journal submissions: elsevier_article
|
841
|
+
* IEEE Transaction journal submissions: ieee_article
|
842
|
+
* JSS articles: jss_article
|
843
|
+
* MDPI journal submissions: mdpi_article
|
844
|
+
* Monthly Notices of the Royal Astronomical Society articles: mnras_article
|
845
|
+
* NNRAS journal submissions: nmras_article
|
846
|
+
* PeerJ articles: peerj_article
|
847
|
+
* Royal Society Open Science journal submissions: rsos_article
|
848
|
+
* Royal Statistical Society: rss_article
|
849
|
+
* Sage journal submissions: sage_article
|
850
|
+
* Springer journal submissions: springer_article
|
851
|
+
* Statistics in Medicine journal submissions: sim_article
|
852
|
+
* Copernicus Publications journal submissions: copernicus_article
|
853
|
+
* The R Journal articles: rjournal_article
|
854
|
+
* Frontiers articles: ???
|
855
|
+
* Taylor & Francis articles: ???
|
856
|
+
* Bulletin De L'AMQ: amq_article
|
857
|
+
* PLOS journal: plos_article
|
858
|
+
* Proceedings of the National Academy of Sciences of the USA: pnas_article
|
859
|
+
|
860
|
+
In order to create a document with one of those templates, use the following command:
|
861
|
+
|
862
|
+
```
|
863
|
+
gknit-draft --filename <my_document> --template <template> --package <package>
|
864
|
+
--create_dir
|
865
|
+
```
|
866
|
+
So, in order to create a template for writing an R Journal, use:
|
867
|
+
|
868
|
+
```
|
869
|
+
gknit-draft --filename my_r_article --template rjournal_article --package rticles
|
870
|
+
--create_dir
|
871
|
+
```
|
872
|
+
|
873
|
+
# Accessing R variables
|
874
|
+
|
875
|
+
Galaaz allows Ruby to access variables created in R. For example, the 'mtcars' data set is
|
876
|
+
available in R and can be accessed from Ruby by using the 'tilda' operator followed by the
|
877
|
+
symbol for the variable, in this case ':mtcar'. In the code bellow method 'outputs' is
|
878
|
+
used to output the 'mtcars' data set nicely formatted in HTML by use of the 'kable' and
|
879
|
+
'kable_styling' functions. Method 'outputs' is only available when used with 'gknit'.
|
880
|
+
|
881
|
+
```{ruby view_kable}
|
882
|
+
outputs (~:mtcars).kable.kable_styling
|
883
|
+
```
|
884
|
+
|
885
|
+
# Basic Data Types
|
886
|
+
|
887
|
+
## Vector
|
99
888
|
|
100
889
|
Vectors can be thought of as contiguous cells containing data. Cells are accessed through
|
101
890
|
indexing operations such as x[5]. Galaaz has six basic (‘atomic’) vector types: logical,
|
@@ -120,20 +909,22 @@ vector is often referred to as a character string.
|
|
120
909
|
To create a vector the 'c' (concatenate) method from the 'R' module should be used:
|
121
910
|
|
122
911
|
```{ruby integer}
|
123
|
-
|
124
|
-
puts
|
912
|
+
vec = R.c(1, 2, 3)
|
913
|
+
puts vec
|
125
914
|
```
|
126
915
|
|
127
|
-
Lets take a look at the type, mode and storage.mode of our vector
|
916
|
+
Lets take a look at the type, mode and storage.mode of our vector vec. In order to print
|
128
917
|
this out, we are creating a data frame 'df' and printing it out. A data frame, for those
|
129
|
-
not familiar with it,
|
918
|
+
not familiar with it, is basically a table. Here we create the data frame and add the
|
130
919
|
column name by passing named parameters for each column, such as 'typeof:', 'mode:' and
|
131
|
-
'storage__mode'. You should also note here that the double underscore is converted to a '.'.
|
920
|
+
'storage__mode?'. You should also note here that the double underscore is converted to a '.'.
|
921
|
+
So, when printed 'storage\_\_mode' will actually print as 'storage.mode'.
|
132
922
|
|
133
|
-
In R, the method used to create a
|
923
|
+
Data frames will later be more carefully described. In R, the method used to create a
|
924
|
+
data frame is 'data.frame', in Galaaz we use 'data\_\_frame'.
|
134
925
|
|
135
926
|
```{ruby typeof_integer}
|
136
|
-
df = R.data__frame(typeof:
|
927
|
+
df = R.data__frame(typeof: vec.typeof, mode: vec.mode, storage__mode: vec.storage__mode)
|
137
928
|
puts df
|
138
929
|
```
|
139
930
|
|
@@ -143,12 +934,12 @@ like '1' is converted to float and to have an integer the R developer will use '
|
|
143
934
|
follows normal Ruby rules and the number 1 is an integer and 1.0 is a float.
|
144
935
|
|
145
936
|
```{ruby float}
|
146
|
-
|
147
|
-
puts
|
937
|
+
vec = R.c(1.0, 2, 3)
|
938
|
+
puts vec
|
148
939
|
```
|
149
940
|
|
150
941
|
```{ruby typeof_float}
|
151
|
-
df = R.data__frame(typeof:
|
942
|
+
df = R.data__frame(typeof: vec.typeof, mode: vec.mode, storage__mode: vec.storage__mode)
|
152
943
|
outputs df.kable.kable_styling
|
153
944
|
```
|
154
945
|
|
@@ -161,47 +952,1101 @@ of the error.
|
|
161
952
|
vec = R.c(1, hello, 5)
|
162
953
|
```
|
163
954
|
|
164
|
-
|
165
|
-
|
955
|
+
Here is a vector with logical values
|
956
|
+
|
957
|
+
```{ruby logical_vector}
|
958
|
+
vec = R.c(true, true, false, false, true)
|
959
|
+
puts vec
|
166
960
|
```
|
167
961
|
|
962
|
+
### Combining Vectors
|
168
963
|
|
169
|
-
|
964
|
+
The 'c' functions used to create vectors can also be used to combine two vectors:
|
170
965
|
|
171
|
-
```{ruby
|
172
|
-
|
966
|
+
```{ruby combining_vectors}
|
967
|
+
vec1 = R.c(10.0, 20.0, 30.0)
|
968
|
+
vec2 = R.c(4.0, 5.0, 6.0)
|
969
|
+
vec = R.c(vec1, vec2)
|
970
|
+
puts vec
|
971
|
+
```
|
972
|
+
In galaaz, methods can be chainned (somewhat like the pipe operator in R %>%, but more generic).
|
973
|
+
In this next example, method 'c' is chainned after 'vec1'. This also looks like 'c' is a
|
974
|
+
method of the vector, but in reallity, this is actually closer to the pipe operator. When
|
975
|
+
Galaaz identifies that 'c' is not a method of 'vec' it actually tries to call 'R.c' with
|
976
|
+
'vec1' as the first argument concatenated with all the other available arguments. The code
|
977
|
+
bellow is automatically converted to the code above.
|
173
978
|
|
174
|
-
|
979
|
+
```{ruby chainning_methods}
|
980
|
+
vec = vec1.c(vec2)
|
981
|
+
puts vec
|
982
|
+
```
|
175
983
|
|
176
|
-
|
177
|
-
mtcars = ~:mtcars
|
178
|
-
mtcars.car_name = R.rownames(:mtcars)
|
179
|
-
# compute normalized mpg
|
180
|
-
mtcars.mpg_z = ((mtcars.mpg - mtcars.mpg.mean)/mtcars.mpg.sd).round 2
|
181
|
-
mtcars.mpg_type = mtcars.mpg_z < 0 ? "below" : "above"
|
182
|
-
mtcars = mtcars[mtcars.mpg_z.order, :all]
|
183
|
-
# convert to factor to retain sorted order in plot
|
184
|
-
mtcars.car_name = mtcars.car_name.factor levels: mtcars.car_name
|
984
|
+
### Vector Arithmetic
|
185
985
|
|
186
|
-
|
187
|
-
gg = mtcars.ggplot(E.aes(x: :car_name, y: :mpg_z, label: :mpg_z)) +
|
188
|
-
R.geom_bar(E.aes(fill: :mpg_type), stat: 'identity', width: 0.5) +
|
189
|
-
R.scale_fill_manual(name: "Mileage",
|
190
|
-
labels: R.c("Above Average", "Below Average"),
|
191
|
-
values: R.c("above": "#00ba38", "below": "#f8766d")) +
|
192
|
-
R.labs(subtitle: "Normalised mileage from 'mtcars'",
|
193
|
-
title: "Diverging Bars") +
|
194
|
-
R.coord_flip()
|
986
|
+
Arithmetic operations on vectors are performed element by element:
|
195
987
|
|
196
|
-
|
988
|
+
```{ruby vec_arith1}
|
989
|
+
puts vec1 + vec2
|
197
990
|
```
|
198
991
|
|
992
|
+
```{ruby mult}
|
993
|
+
puts vec1 * 5
|
994
|
+
```
|
199
995
|
|
200
|
-
|
996
|
+
When vectors have different length, a recycling rule is applied to the shorter vector:
|
201
997
|
|
998
|
+
```{ruby recycle}
|
999
|
+
vec3 = R.c(1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0)
|
1000
|
+
puts vec4 = vec1 + vec3
|
1001
|
+
```
|
202
1002
|
|
203
|
-
|
1003
|
+
### Vector Indexing
|
1004
|
+
|
1005
|
+
Vectors can be indexed by using the '[]' operator:
|
1006
|
+
|
1007
|
+
```{ruby index}
|
1008
|
+
puts vec4[3]
|
1009
|
+
```
|
1010
|
+
|
1011
|
+
We can also index a vector with another vector. For example, in the code bellow, we take elements
|
1012
|
+
1, 3, 5, and 7 from vec3:
|
1013
|
+
|
1014
|
+
```{ruby index_by_vector}
|
1015
|
+
puts vec4[R.c(1, 3, 5, 7)]
|
1016
|
+
```
|
1017
|
+
|
1018
|
+
Repeating an index and having indices out of order is valid code:
|
1019
|
+
|
1020
|
+
```{ruby repeated_index}
|
1021
|
+
puts vec4[R.c(1, 3, 3, 1)]
|
1022
|
+
```
|
1023
|
+
|
1024
|
+
It is also possible to index a vector with a negative number or negative vector. In these cases
|
1025
|
+
the indexed values are not returned:
|
1026
|
+
|
1027
|
+
```{ruby neg_index}
|
1028
|
+
puts vec4[-3]
|
1029
|
+
puts vec4[-R.c(1, 3, 5, 7)]
|
1030
|
+
```
|
1031
|
+
|
1032
|
+
If an index is out of range, a missing value (NA) will be reported.
|
1033
|
+
|
1034
|
+
```{ruby out_of_range}
|
1035
|
+
puts vec4[30]
|
1036
|
+
```
|
1037
|
+
|
1038
|
+
It is also possible to index a vector by range:
|
1039
|
+
|
1040
|
+
```{ruby range}
|
1041
|
+
puts vec4[(2..5)]
|
1042
|
+
```
|
1043
|
+
|
1044
|
+
Elements in a vector can be named using the 'names' attribute of a vector:
|
1045
|
+
|
1046
|
+
```{ruby naming}
|
1047
|
+
full_name = R.c("Rodrigo", "A", "Botafogo")
|
1048
|
+
full_name.names = R.c("First", "Middle", "Last")
|
1049
|
+
puts full_name
|
1050
|
+
```
|
1051
|
+
|
1052
|
+
Or it can also be named by using the 'c' function with named paramenters:
|
1053
|
+
|
1054
|
+
```{ruby named_param}
|
1055
|
+
full_name = R.c(First: "Rodrigo", Middle: "A", Last: "Botafogo")
|
1056
|
+
puts full_name
|
1057
|
+
```
|
1058
|
+
|
1059
|
+
### Extracting Native Ruby Types from a Vector
|
1060
|
+
|
1061
|
+
Vectors created with 'R.c' are of class R::Vector. You might have noticed that when indexing a
|
1062
|
+
vector, a new vector is returned, even if this vector has one single element. In order to use
|
1063
|
+
R::Vector with other ruby classes it might be necessary to extract the actual Ruby native type
|
1064
|
+
from the vector. In order to do this extraction the '>>' operator is used.
|
1065
|
+
|
1066
|
+
```{ruby ruby_native}
|
1067
|
+
puts vec4
|
1068
|
+
puts vec4 >> 0
|
1069
|
+
puts vec4 >> 4
|
1070
|
+
```
|
1071
|
+
|
1072
|
+
Note that indexing with '>>' starts at 0 and not at 1, also, we cannot do negative indexing.
|
1073
|
+
|
1074
|
+
## Matrix
|
1075
|
+
|
1076
|
+
A matrix is a collection of elements organized as a two dimensional table. A matrix can be
|
1077
|
+
created by the 'matrix' function:
|
1078
|
+
|
1079
|
+
```{ruby matrix}
|
1080
|
+
mat = R.matrix(R.c(1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0),
|
1081
|
+
nrow: 3,
|
1082
|
+
ncol: 3)
|
1083
|
+
|
1084
|
+
puts mat
|
1085
|
+
```
|
1086
|
+
Note that matrices data is organized by column first. It is possible to organize the matrix
|
1087
|
+
memory by row first passing an extra argument to the 'matrix' function:
|
1088
|
+
|
1089
|
+
```{ruby matrix_rowfirst}
|
1090
|
+
mat_row = R.matrix(R.c(1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0),
|
1091
|
+
nrow: 3,
|
1092
|
+
ncol: 3,
|
1093
|
+
byrow: true)
|
1094
|
+
|
1095
|
+
puts mat_row
|
1096
|
+
```
|
1097
|
+
|
1098
|
+
### Indexing a Matrix
|
1099
|
+
|
1100
|
+
A matrix can be indexed by [row, column]:
|
1101
|
+
|
1102
|
+
```{ruby matrix_index}
|
1103
|
+
puts mat_row[1, 1]
|
1104
|
+
puts mat_row[2, 3]
|
1105
|
+
```
|
1106
|
+
It is possible to index an entire row or column with the ':all' keyword
|
1107
|
+
|
1108
|
+
```{ruby matrix_index_all}
|
1109
|
+
puts mat_row[1, :all]
|
1110
|
+
puts mat_row[:all, 2]
|
1111
|
+
```
|
1112
|
+
|
1113
|
+
Indexing with a vector is also possible for matrices. In the following example we want
|
1114
|
+
rows 1 and 3 and columns 2 and 3 building a 2 x 2 matrix.
|
1115
|
+
|
1116
|
+
```{ruby matrix_index_vector}
|
1117
|
+
puts mat_row[R.c(1, 3), R.c(2, 3)]
|
1118
|
+
```
|
1119
|
+
|
1120
|
+
Matrices can be combined with functions 'rbind':
|
1121
|
+
|
1122
|
+
```{ruby matrix_combine_rbind}
|
1123
|
+
puts mat_row.rbind(mat)
|
1124
|
+
```
|
1125
|
+
|
1126
|
+
and 'cbind':
|
1127
|
+
|
1128
|
+
```{ruby matrix_combine_cbind}
|
1129
|
+
puts mat_row.cbind(mat)
|
1130
|
+
```
|
1131
|
+
|
1132
|
+
## List
|
1133
|
+
|
1134
|
+
A list is a data structure that can contain sublists of different types, while vector and matrix
|
1135
|
+
can only hold one type of element.
|
1136
|
+
|
1137
|
+
```{ruby list}
|
1138
|
+
nums = R.c(1.0, 2.0, 3.0)
|
1139
|
+
strs = R.c("a", "b", "c", "d")
|
1140
|
+
bool = R.c(true, true, false)
|
1141
|
+
lst = R.list(nums: nums, strs: strs, bool: bool)
|
1142
|
+
puts lst
|
1143
|
+
```
|
1144
|
+
|
1145
|
+
Note that 'lst' elements are named elements.
|
1146
|
+
|
1147
|
+
|
1148
|
+
### List Indexing
|
1149
|
+
|
1150
|
+
List indexing, also called slicing, is done using the '[]' operator and the '[[]]' operator. Let's
|
1151
|
+
first start with the '[]' operator. The list above has three sublist indexing with '[]' will
|
1152
|
+
return one of the sublists.
|
1153
|
+
|
1154
|
+
```{ruby list_indexing}
|
1155
|
+
puts lst[1]
|
1156
|
+
```
|
1157
|
+
|
1158
|
+
Note that when using '[]' a new list is returned. When using the double square bracket operator
|
1159
|
+
the value returned is the actual element of the list in the given position and not a slice of
|
1160
|
+
the original list
|
1161
|
+
|
1162
|
+
|
1163
|
+
```{ruby list_indexing_single}
|
1164
|
+
puts lst[[1]]
|
1165
|
+
```
|
1166
|
+
|
1167
|
+
When elements are named, as dones with lst, indexing can be done by name:
|
1168
|
+
|
1169
|
+
```{ruby list_indexing_by_name}
|
1170
|
+
puts lst[['bool']][[1]] >> 0
|
1171
|
+
```
|
1172
|
+
|
1173
|
+
In this example, first the 'bool' element of the list was extracted, not as a list, but as a vector,
|
1174
|
+
then the first element of the vector was extracted (note that vectors also accept the '[[]]'
|
1175
|
+
operator) and then the vector was indexed by its first element, extracting the native Ruby type.
|
1176
|
+
|
1177
|
+
|
1178
|
+
## Data Frame
|
1179
|
+
|
1180
|
+
A data frame is a table like structure in which each column has the same number of
|
1181
|
+
rows. Data frames are the basic structure for storing data for data analysis. We have already
|
1182
|
+
seen a data frame previously when we accessed variable '~:mtcars'. In order to create a
|
1183
|
+
data frame, function 'data__frame' is used:
|
1184
|
+
|
1185
|
+
```{ruby dataframe}
|
1186
|
+
df = R.data__frame(
|
1187
|
+
year: R.c(2010, 2011, 2012),
|
1188
|
+
income: R.c(1000.0, 1500.0, 2000.0))
|
1189
|
+
|
1190
|
+
puts df
|
1191
|
+
```
|
1192
|
+
|
1193
|
+
### Data Frame Indexing
|
1194
|
+
|
1195
|
+
A data frame can be indexed the same way as a matrix, by using '[row, column]', where row and
|
1196
|
+
column can either be a numeric or the name of the row or column
|
1197
|
+
|
1198
|
+
```{ruby dataframe_index}
|
1199
|
+
puts (~:mtcars).head
|
1200
|
+
puts (~:mtcars)[1, 2]
|
1201
|
+
puts (~:mtcars)['Datsun 710', 'mpg']
|
1202
|
+
```
|
1203
|
+
|
1204
|
+
Extracting a column from a data frame as a vector can be done by using the double square bracket
|
1205
|
+
operator:
|
1206
|
+
|
1207
|
+
```{ruby dataframe_column}
|
1208
|
+
puts (~:mtcars)[['mpg']]
|
1209
|
+
```
|
1210
|
+
|
1211
|
+
A data frame column can also be accessed as if it were an instance variable of the data frame:
|
1212
|
+
|
1213
|
+
```{ruby dataframe_instance_variable}
|
1214
|
+
puts (~:mtcars).mpg
|
1215
|
+
```
|
1216
|
+
|
1217
|
+
Slicing a data frame can be done by indexing it with a vector (we use 'head' to reduce the
|
1218
|
+
output):
|
1219
|
+
|
1220
|
+
```{ruby dataframe_column_slice}
|
1221
|
+
puts (~:mtcars)[R.c('mpg', 'hp')].head
|
1222
|
+
```
|
1223
|
+
|
1224
|
+
A row slice can be obtained by indexing by row and using the ':all' keyword for the column:
|
1225
|
+
|
1226
|
+
```{ruby dataframe_row_slice}
|
1227
|
+
puts (~:mtcars)[R.c('Datsun 710', 'Camaro Z28'), :all]
|
1228
|
+
```
|
1229
|
+
|
1230
|
+
Finally, a data frame can also be indexed with a logical vector. In this next example, the
|
1231
|
+
'am' column of :mtcars is compared with 0 (with method 'eq'). When 'am' is equal to 0 the
|
1232
|
+
car is automatic. So, by doing '(~:mtcars).am.eq 0' a logical vector is created with
|
1233
|
+
'true' whenever 'am' is 0 and 'false' otherwise.
|
1234
|
+
|
1235
|
+
```{ruby logical_vector_filter}
|
1236
|
+
# obtain a vector with 'true' for cars with automatic transmission
|
1237
|
+
automatic = (~:mtcars).am.eq 0
|
1238
|
+
puts automatic
|
1239
|
+
```
|
1240
|
+
|
1241
|
+
Using this logical vector, the data frame is indexed, returning a new data frame in
|
1242
|
+
which all cars have automatic transmission.
|
1243
|
+
|
1244
|
+
```{ruby dataframe_logical}
|
1245
|
+
# slice the data frame by using this vector
|
1246
|
+
puts (~:mtcars)[automatic, :all]
|
1247
|
+
```
|
1248
|
+
|
1249
|
+
# Writing Expressions in Galaaz
|
1250
|
+
|
1251
|
+
Galaaz extends Ruby to work with complex expressions, similar to R's expressions build with 'quote'
|
1252
|
+
(base R) or 'quo' (tidyverse). Let's take a look at some of those expressions.
|
1253
|
+
|
1254
|
+
## Expressions from operators
|
1255
|
+
|
1256
|
+
The code bellow
|
1257
|
+
creates an expression summing two symbols
|
1258
|
+
|
1259
|
+
```{ruby expressions}
|
1260
|
+
exp1 = :a + :b
|
1261
|
+
puts exp1
|
1262
|
+
```
|
1263
|
+
We can build any complex mathematical expression
|
1264
|
+
|
1265
|
+
```{ruby expr2}
|
1266
|
+
exp2 = (:a + :b) * 2.0 + :c ** 2 / :z
|
1267
|
+
puts exp2
|
1268
|
+
```
|
1269
|
+
|
1270
|
+
It is also possible to use inequality operators in building expressions
|
1271
|
+
|
1272
|
+
```{ruby expr3}
|
1273
|
+
exp3 = (:a + :b) >= :z
|
1274
|
+
puts exp3
|
1275
|
+
```
|
1276
|
+
|
1277
|
+
Galaaz provides both symbolic representations for operators, such as (>, <, !=) as functional
|
1278
|
+
notation for those operators such as (.gt, .ge, etc.). So the same expression written
|
1279
|
+
above can also be written as
|
1280
|
+
|
1281
|
+
```{ruby expr4}
|
1282
|
+
exp4 = (:a + :b).ge :z
|
1283
|
+
puts exp4
|
1284
|
+
```
|
1285
|
+
|
1286
|
+
Two type of expression can only be created with the functional representation of the operators,
|
1287
|
+
those are expressions involving '==', and '='. In order to write an expression involving '==' we
|
1288
|
+
need to use the method '.eq' and for '=' we need the function '.assign'
|
1289
|
+
|
1290
|
+
```{ruby expr5}
|
1291
|
+
exp5 = (:a + :b).eq :z
|
1292
|
+
puts exp5
|
1293
|
+
```
|
1294
|
+
|
1295
|
+
```{ruby expr6}
|
1296
|
+
exp6 = :y.assign :a + :b
|
1297
|
+
puts exp6
|
1298
|
+
```
|
1299
|
+
In general we think that using the functional notation is preferable to using the
|
1300
|
+
symbolic notation as otherwise, we end up writing invalid expressions such as
|
1301
|
+
|
1302
|
+
```{ruby exp_wrong, warning=FALSE, eval=FALSE}
|
1303
|
+
exp_wrong = (:a + :b) == :z
|
1304
|
+
puts exp_wrong
|
1305
|
+
```
|
1306
|
+
and it might be difficult to understand what is going on here. The problem lies with the fact that
|
1307
|
+
when using '==' we are comparing expression (:a + :b) to expression :z with '=='. When the
|
1308
|
+
comparison is executed, the system tries to evaluate :a, :b and :z, and those symbols at
|
1309
|
+
this time are not bound to anything and we get a "object 'a' not found" message.
|
1310
|
+
If we only use functional notation, this type of error will not occur.
|
1311
|
+
|
1312
|
+
## Expressions with R methods
|
1313
|
+
|
1314
|
+
It is often necessary to create an expression that uses a method or function. For instance, in
|
1315
|
+
mathematics, it's quite natural to write an expressin such as $y = sin(x)$. In this case, the
|
1316
|
+
'sin' function is part of the expression and should not immediately executed. Now, let's say
|
1317
|
+
that 'x' is an angle of 45$^\circ$ and we acttually want our expression to be $y = 0.850...$.
|
1318
|
+
When we want the function to be part of the expression, we call the function preceeding it
|
1319
|
+
by the letter E, such as 'E.sin(x)'
|
1320
|
+
|
1321
|
+
```{ruby method_expression}
|
1322
|
+
exp7 = :y.assign E.sin(:x)
|
1323
|
+
puts exp7
|
1324
|
+
```
|
1325
|
+
|
1326
|
+
Expressions can also be written using '.' notation:
|
1327
|
+
|
1328
|
+
```{ruby expression_with_dot}
|
1329
|
+
exp8 = :y.assign :x.sin
|
1330
|
+
puts exp8
|
1331
|
+
```
|
1332
|
+
|
1333
|
+
When a function has multiple arguments, the first one can be used before the '.':
|
1334
|
+
|
1335
|
+
```{ruby expression_multiple_args}
|
1336
|
+
exp9 = :x.c(:y)
|
1337
|
+
puts exp9
|
1338
|
+
```
|
1339
|
+
|
1340
|
+
## Evaluating an Expression
|
1341
|
+
|
1342
|
+
Expressions can be evaluated by calling function 'eval' with a binding. A binding can be provided
|
1343
|
+
with a list:
|
1344
|
+
|
1345
|
+
```{ruby eval_expression_list}
|
1346
|
+
exp = (:a + :b) * 2.0 + :c ** 2 / :z
|
1347
|
+
puts exp.eval(R.list(a: 10, b: 20, c: 30, z: 40))
|
1348
|
+
```
|
1349
|
+
|
1350
|
+
... with a data frame:
|
1351
|
+
|
1352
|
+
```{ruby eval_expression_df}
|
1353
|
+
df = R.data__frame(
|
1354
|
+
a: R.c(1, 2, 3),
|
1355
|
+
b: R.c(10, 20, 30),
|
1356
|
+
c: R.c(100, 200, 300),
|
1357
|
+
z: R.c(1000, 2000, 3000))
|
1358
|
+
|
1359
|
+
puts exp.eval(df)
|
1360
|
+
```
|
1361
|
+
|
1362
|
+
# Manipulating Data
|
1363
|
+
|
1364
|
+
One of the major benefits of Galaaz is to bring strong data manipulation to Ruby. The following
|
1365
|
+
examples were extracted from Hardley's "R for Data Science" (https://r4ds.had.co.nz/). This
|
1366
|
+
is a highly recommended book for those not already familiar with the 'tidyverse' style of
|
1367
|
+
programming in R. In the sections to follow, we will limit ourselves to convert the R code to
|
1368
|
+
Galaaz.
|
1369
|
+
|
1370
|
+
For these
|
1371
|
+
examples, we will investigate the nycflights13 data set available on the package by the
|
1372
|
+
same name. We use function 'R.install\_and\_loads' that checks if the library is available
|
1373
|
+
locally, and if not, installs it. This data frame contains all 336,776 flights that
|
1374
|
+
departed from New York City in 2013. The data comes from the US Bureau of
|
1375
|
+
Transportation Statistics.
|
1376
|
+
|
1377
|
+
Dplyr uses 'tibbles' in place of data frames; unfortunately, tibbles do not print yet properly in
|
1378
|
+
Galaaz due to a bug in fastR. In order to print a tibble we need to convert it to a data frame
|
1379
|
+
using the 'as\_\_data__frame' method.
|
1380
|
+
|
1381
|
+
```{ruby nycflights13}
|
1382
|
+
R.install_and_loads('nycflights13')
|
1383
|
+
R.library('dplyr')
|
1384
|
+
```
|
1385
|
+
|
1386
|
+
```{ruby flights}
|
1387
|
+
flights = ~:flights
|
1388
|
+
puts flights.head
|
1389
|
+
```
|
1390
|
+
|
1391
|
+
## Filtering rows with Filter
|
1392
|
+
|
1393
|
+
In this example we filter the flights data set by giving to the filter function two expressions:
|
1394
|
+
the first :month.eq 1
|
1395
|
+
|
1396
|
+
```{ruby filter_rows}
|
1397
|
+
puts flights.filter((:month.eq 1), (:day.eq 1)).head
|
1398
|
+
```
|
1399
|
+
|
1400
|
+
## Logical Operators
|
1401
|
+
|
1402
|
+
All flights that departed in November of December
|
1403
|
+
|
1404
|
+
```{ruby nov_dec}
|
1405
|
+
puts flights.filter((:month.eq 11) | (:month.eq 12)).head
|
1406
|
+
```
|
1407
|
+
|
1408
|
+
The same as above, but using the 'in' operator. In R, it is possible to define many operators
|
1409
|
+
by doing %<op>%. The %in% operator checks if a value is in a vector. In order to use those
|
1410
|
+
operators from Galaaz the '._' method is used, where the first argument is the operator's
|
1411
|
+
symbol, in this case ':in' and the second argument is the vector:
|
1412
|
+
|
1413
|
+
```{ruby in_op}
|
1414
|
+
puts flights.filter(:month._ :in, R.c(11, 12)).head
|
1415
|
+
```
|
1416
|
+
|
1417
|
+
## Filtering with NA (Not Available)
|
1418
|
+
|
1419
|
+
Let's first create a 'tibble' with a Not Available value (R::NA). Tibbles are a modern
|
1420
|
+
version of a data frame and operate very similarly to one. It differs in how it outputs
|
1421
|
+
the values and the result of some subsetting operations that are more consistent than
|
1422
|
+
what is obtained from data frame.
|
1423
|
+
|
1424
|
+
```{ruby na_tibble}
|
1425
|
+
df = R.tibble(x: R.c(1, R::NA, 3))
|
1426
|
+
puts df
|
1427
|
+
```
|
1428
|
+
|
1429
|
+
Now filtering by :x > 1 shows all lines that satisfy this condition, where the row with R:NA does
|
1430
|
+
not.
|
1431
|
+
|
1432
|
+
```{ruby filter_na}
|
1433
|
+
puts df.filter(:x > 1)
|
1434
|
+
```
|
1435
|
+
|
1436
|
+
To match an NA use method 'is__na'
|
1437
|
+
|
1438
|
+
```{ruby with_na}
|
1439
|
+
puts df.filter((:x.is__na) | (:x > 1))
|
1440
|
+
```
|
1441
|
+
|
1442
|
+
## Arrange Rows with arrange
|
1443
|
+
|
1444
|
+
Arrange reorders the rows of a data frame by the given arguments.
|
1445
|
+
|
1446
|
+
```{ruby arrange}
|
1447
|
+
puts flights.arrange(:year, :month, :day).head
|
1448
|
+
```
|
1449
|
+
|
1450
|
+
To arrange in descending order, use function 'desc'
|
1451
|
+
|
1452
|
+
```{ruby desc_arrange}
|
1453
|
+
puts flights.arrange(:dep_delay.desc).head
|
1454
|
+
```
|
1455
|
+
|
1456
|
+
## Selecting columns
|
1457
|
+
|
1458
|
+
To select specific columns from a dataset we use function 'select':
|
1459
|
+
|
1460
|
+
```{ruby select}
|
1461
|
+
puts flights.select(:year, :month, :day).head
|
1462
|
+
```
|
1463
|
+
|
1464
|
+
It is also possible to select column in a given range
|
1465
|
+
|
1466
|
+
```{ruby select_range}
|
1467
|
+
puts flights.select(:year.up_to :day).head
|
1468
|
+
```
|
1469
|
+
|
1470
|
+
Select all columns that start with a given name sequence
|
204
1471
|
|
1472
|
+
```{ruby select_starts_with}
|
1473
|
+
puts flights.select(E.starts_with('arr')).head
|
1474
|
+
```
|
1475
|
+
|
1476
|
+
Other functions that can be used:
|
1477
|
+
|
1478
|
+
* ends_with("xyz"): matches names that end with “xyz”.
|
1479
|
+
|
1480
|
+
* contains("ijk"): matches names that contain “ijk”.
|
1481
|
+
|
1482
|
+
* matches("(.)\\1"): selects variables that match a regular expression. This one matches
|
1483
|
+
any variables that contain repeated characters.
|
1484
|
+
|
1485
|
+
* num_range("x", (1..3)): matches x1, x2 and x3
|
1486
|
+
|
1487
|
+
A helper function that comes in handy when we just want to rearrange column order is 'Everything':
|
1488
|
+
|
1489
|
+
```{ruby everything}
|
1490
|
+
puts flights.select(:year, :month, :day, E.everything).head
|
1491
|
+
```
|
1492
|
+
|
1493
|
+
## Add variables to a dataframe with 'mutate'
|
1494
|
+
|
1495
|
+
```{ruby small_flights}
|
1496
|
+
flights_sm = flights.
|
1497
|
+
select((:year.up_to :day),
|
1498
|
+
E.ends_with('delay'),
|
1499
|
+
:distance,
|
1500
|
+
:air_time)
|
1501
|
+
|
1502
|
+
puts flights_sm.head
|
1503
|
+
```
|
1504
|
+
|
1505
|
+
```{ruby mutate}
|
1506
|
+
flights_sm = flights_sm.
|
1507
|
+
mutate(gain: :dep_delay - :arr_delay,
|
1508
|
+
speed: :distance / :air_time * 60)
|
1509
|
+
puts flights_sm.head
|
1510
|
+
```
|
1511
|
+
|
1512
|
+
## Summarising data
|
1513
|
+
|
1514
|
+
Function 'summarise' calculates summaries for the data frame. When no 'group_by' is used
|
1515
|
+
a single value is obtained from the data frame:
|
1516
|
+
|
1517
|
+
```{ruby summarise}
|
1518
|
+
puts flights.summarise(delay: E.mean(:dep_delay, na__rm: true))
|
1519
|
+
```
|
1520
|
+
|
1521
|
+
When a data frame is grouped with 'group_by' summaries apply to the given group:
|
1522
|
+
|
1523
|
+
```{ruby summarise_group_by}
|
1524
|
+
by_day = flights.group_by(:year, :month, :day)
|
1525
|
+
puts by_day.summarise(delay: :dep_delay.mean(na__rm: true)).head
|
1526
|
+
```
|
1527
|
+
|
1528
|
+
Next we put many operations together by pipping them one after the other:
|
1529
|
+
|
1530
|
+
```{ruby pipping}
|
1531
|
+
delays = flights.
|
1532
|
+
group_by(:dest).
|
1533
|
+
summarise(
|
1534
|
+
count: E.n,
|
1535
|
+
dist: :distance.mean(na__rm: true),
|
1536
|
+
delay: :arr_delay.mean(na__rm: true)).
|
1537
|
+
filter(:count > 20, :dest != "NHL")
|
1538
|
+
|
1539
|
+
puts delays.head
|
1540
|
+
```
|
1541
|
+
|
1542
|
+
# Using Data Table
|
1543
|
+
|
1544
|
+
```{ruby fread}
|
1545
|
+
R.library('data.table')
|
1546
|
+
R.install_and_loads('curl')
|
1547
|
+
|
1548
|
+
input = "https://raw.githubusercontent.com/Rdatatable/data.table/master/vignettes/flights14.csv"
|
1549
|
+
flights = R.fread(input)
|
1550
|
+
puts flights
|
1551
|
+
puts flights.dim
|
1552
|
+
```
|
1553
|
+
|
1554
|
+
```{ruby data_table}
|
1555
|
+
|
1556
|
+
data_table = R.data__table(
|
1557
|
+
ID: R.c("b","b","b","a","a","c"),
|
1558
|
+
a: (1..6),
|
1559
|
+
b: (7..12),
|
1560
|
+
c: (13..18)
|
1561
|
+
)
|
1562
|
+
|
1563
|
+
puts data_table
|
1564
|
+
puts data_table.ID
|
1565
|
+
```
|
1566
|
+
|
1567
|
+
```{ruby subset_i}
|
1568
|
+
# subset rows in i
|
1569
|
+
ans = flights[(:origin.eq "JFK") & (:month.eq 6)]
|
1570
|
+
puts ans.head
|
1571
|
+
|
1572
|
+
# Get the first two rows from flights.
|
1573
|
+
|
1574
|
+
ans = flights[(1..2)]
|
1575
|
+
puts ans
|
1576
|
+
|
1577
|
+
# Sort flights first by column origin in ascending order, and then by dest in descending order:
|
1578
|
+
|
1579
|
+
# ans = flights[E.order(:origin, -(:dest))]
|
1580
|
+
# puts ans.head
|
1581
|
+
|
1582
|
+
```
|
1583
|
+
|
1584
|
+
```{ruby select_j}
|
1585
|
+
# Select column(s) in j
|
1586
|
+
# select arr_delay column, but return it as a vector.
|
1587
|
+
|
1588
|
+
ans = flights[:all, :arr_delay]
|
1589
|
+
puts ans.head
|
1590
|
+
|
1591
|
+
# Select arr_delay column, but return as a data.table instead.
|
1592
|
+
|
1593
|
+
ans = flights[:all, :arr_delay.list]
|
1594
|
+
puts ans.head
|
1595
|
+
|
1596
|
+
ans = flights[:all, E.list(:arr_delay, :dep_delay)]
|
1597
|
+
```
|
1598
|
+
|
1599
|
+
# Graphics in Galaaz
|
1600
|
+
|
1601
|
+
Creating graphics in Galaaz is quite easy, as it can use all the power of ggplot2. There are
|
1602
|
+
many resources in the web that teaches ggplot, so here we give a quick example of ggplot
|
1603
|
+
integration with Ruby. We continue to use the :mtcars dataset and we will plot a diverging
|
1604
|
+
bar plot, showing cars that have 'above' or 'below' gas consuption. Let's first prepare
|
1605
|
+
the data frame with the necessary data:
|
1606
|
+
|
1607
|
+
```{ruby diverging_plot_pre}
|
1608
|
+
# copy the R variable :mtcars to the Ruby mtcars variable
|
1609
|
+
mtcars = ~:mtcars
|
1610
|
+
|
1611
|
+
# create a new column 'car_name' to store the car names so that it can be
|
1612
|
+
# used for plotting. The 'rownames' of the data frame cannot be used as
|
1613
|
+
# data for plotting
|
1614
|
+
mtcars.car_name = R.rownames(:mtcars)
|
1615
|
+
|
1616
|
+
# compute normalized mpg and add it to a new column called mpg_z
|
1617
|
+
# Note that the mean value for mpg can be obtained by calling the 'mean'
|
1618
|
+
# function on the vector 'mtcars.mpg'. The same with the standard
|
1619
|
+
# deviation 'sd'. The vector is then rounded to two digits with 'round 2'
|
1620
|
+
mtcars.mpg_z = ((mtcars.mpg - mtcars.mpg.mean)/mtcars.mpg.sd).round 2
|
1621
|
+
|
1622
|
+
# create a new column 'mpg_type'. Function 'ifelse' is a vectorized function
|
1623
|
+
# that looks at every element of the mpg_z vector and if the value is below
|
1624
|
+
# 0, returns 'below', otherwise returns 'above'
|
1625
|
+
mtcars.mpg_type = (mtcars.mpg_z < 0).ifelse("below", "above")
|
1626
|
+
|
1627
|
+
# order the mtcar data set by the mpg_z vector from smaler to larger values
|
1628
|
+
mtcars = mtcars[mtcars.mpg_z.order, :all]
|
1629
|
+
|
1630
|
+
# convert the car_name column to a factor to retain sorted order in plot
|
1631
|
+
mtcars.car_name = mtcars.car_name.factor levels: mtcars.car_name
|
1632
|
+
|
1633
|
+
# let's look at the final data frame
|
1634
|
+
puts mtcars.head
|
1635
|
+
```
|
1636
|
+
Now, lets plot the diverging bar plot. When using gKnit, there is no need to call
|
1637
|
+
'R.awt' to create a plotting device, since gKnit does take care of it. Galaaz
|
1638
|
+
provides integration with ggplot. The interested reader should check online for more
|
1639
|
+
information on ggplot, since it is outside the scope of this manual describing
|
1640
|
+
how ggplot works. We give here but a brief description on how this plot is generated.
|
1641
|
+
|
1642
|
+
ggplot implements the 'grammar of graphics'. In this approach, plots are build by
|
1643
|
+
adding layers to the plot. On the first layer we describe what we want on the 'x'
|
1644
|
+
and 'y' axis of the plot. In this case, we have 'car_name' on the 'x' axis and
|
1645
|
+
'mpg\_z' on the 'y' axis. Then the type of graph is specified by adding
|
1646
|
+
'geom\_bar' (for a bar graph). We specify that our bars should be filled using
|
1647
|
+
'mpg\_type', which is either 'above' or 'bellow' giving then two colours for
|
1648
|
+
filling. On the next layer we specify the labels for the graph, then we add the
|
1649
|
+
title and subtitle. Finally, in a bar chart usually bars go on the vertical direction,
|
1650
|
+
but in this graph we want the bars to be horizontally layed so we add 'coord\_flip'.
|
1651
|
+
|
1652
|
+
```{ruby diverging_bar, fig.width = 9.1, fig.height = 6.5}
|
1653
|
+
require 'ggplot'
|
1654
|
+
|
1655
|
+
puts mtcars.ggplot(E.aes(x: :car_name, y: :mpg_z, label: :mpg_z)) +
|
1656
|
+
R.geom_bar(E.aes(fill: :mpg_type), stat: 'identity', width: 0.5) +
|
1657
|
+
R.scale_fill_manual(name: 'Mileage',
|
1658
|
+
labels: R.c('Above Average', 'Below Average'),
|
1659
|
+
values: R.c('above': '#00ba38', 'below': '#f8766d')) +
|
1660
|
+
R.labs(subtitle: "Normalised mileage from 'mtcars'",
|
1661
|
+
title: "Diverging Bars") +
|
1662
|
+
R.coord_flip
|
1663
|
+
```
|
1664
|
+
|
1665
|
+
# Coding with Tidyverse
|
1666
|
+
|
1667
|
+
In R, and when coding with 'tidyverse', arguments to a function are usually not
|
1668
|
+
*referencially transparent*. That is, you can’t replace a value with a seemingly equivalent
|
1669
|
+
object that you’ve defined elsewhere. To see the problem, let's first define a data frame:
|
1670
|
+
|
1671
|
+
```{ruby df}
|
1672
|
+
df = R.data__frame(x: (1..3), y: (3..1))
|
1673
|
+
puts df
|
1674
|
+
```
|
1675
|
+
|
1676
|
+
and now, let's look at this code:
|
1677
|
+
|
1678
|
+
```{r not_transp, eval=FALSE}
|
1679
|
+
my_var <- x
|
1680
|
+
filter(df, my_var == 1)
|
1681
|
+
```
|
1682
|
+
It generates the following error: "object 'x' not found.
|
1683
|
+
|
1684
|
+
However, in Galaaz, arguments are referencially transparent as can be seen by the
|
1685
|
+
code bellow. Note initally that 'my_var = :x' will not give the error "object 'x' not found"
|
1686
|
+
since ':x' is treated as an expression and assigned to my\_var. Then when doing (my\_var.eq 1),
|
1687
|
+
my\_var is a variable that resolves to ':x' and it becomes equivalent to (:x.eq 1) which is
|
1688
|
+
what we want.
|
1689
|
+
|
1690
|
+
```{ruby my_var}
|
1691
|
+
my_var = :x
|
1692
|
+
puts df.filter(my_var.eq 1)
|
1693
|
+
```
|
1694
|
+
As stated by Hardley
|
1695
|
+
|
1696
|
+
> dplyr code is ambiguous. Depending on what variables are defined where,
|
1697
|
+
> filter(df, x == y) could be equivalent to any of:
|
1698
|
+
|
1699
|
+
```
|
1700
|
+
df[df$x == df$y, ]
|
1701
|
+
df[df$x == y, ]
|
1702
|
+
df[x == df$y, ]
|
1703
|
+
df[x == y, ]
|
1704
|
+
```
|
1705
|
+
In galaaz this ambiguity does not exist, filter(df, x.eq y) is not a valid expression as
|
1706
|
+
expressions are build with symbols. In doing filter(df, :x.eq y) we are looking for elements
|
1707
|
+
of the 'x' column that are equal to a previously defined y variable. Finally in
|
1708
|
+
filter(df, :x.eq :y) we are looking for elements in which the 'x' column value is equal to
|
1709
|
+
the 'y' column value. This can be seen in the following two chunks of code:
|
1710
|
+
|
1711
|
+
```{ruby disamb1}
|
1712
|
+
y = 1
|
1713
|
+
x = 2
|
1714
|
+
|
1715
|
+
# looking for values where the 'x' column is equal to the 'y' column
|
1716
|
+
puts df.filter(:x.eq :y)
|
1717
|
+
```
|
1718
|
+
|
1719
|
+
```{ruby disamb2}
|
1720
|
+
# looking for values where the 'x' column is equal to the 'y' variable
|
1721
|
+
# in this case, the number 1
|
1722
|
+
puts df.filter(:x.eq y)
|
1723
|
+
```
|
1724
|
+
## Writing a function that applies to different data sets
|
1725
|
+
|
1726
|
+
Let's suppose that we want to write a function that receives as the first argument a data frame
|
1727
|
+
and as second argument an expression that adds a column to the data frame that is equal to the
|
1728
|
+
sum of elements in column 'a' plus 'x'.
|
1729
|
+
|
1730
|
+
Here is the intended behaviour using the 'mutate' function of 'dplyr':
|
1731
|
+
|
1732
|
+
```
|
1733
|
+
mutate(df1, y = a + x)
|
1734
|
+
mutate(df2, y = a + x)
|
1735
|
+
mutate(df3, y = a + x)
|
1736
|
+
mutate(df4, y = a + x)
|
1737
|
+
```
|
1738
|
+
The naive approach to writing an R function to solve this problem is:
|
1739
|
+
|
1740
|
+
```
|
1741
|
+
mutate_y <- function(df) {
|
1742
|
+
mutate(df, y = a + x)
|
1743
|
+
}
|
1744
|
+
```
|
1745
|
+
Unfortunately, in R, this function can fail silently if one of the variables isn’t present
|
1746
|
+
in the data frame, but is present in the global environment. We will not go through here how
|
1747
|
+
to solve this problem in R.
|
1748
|
+
|
1749
|
+
In Galaaz the method mutate_y bellow will work fine and will never fail silently.
|
1750
|
+
|
1751
|
+
```{ruby mutate_y, warning=FALSE}
|
1752
|
+
def mutate_y(df)
|
1753
|
+
df.mutate(:y.assign :a + :x)
|
1754
|
+
end
|
1755
|
+
```
|
1756
|
+
Here we create a data frame that has only one column named 'x':
|
1757
|
+
|
1758
|
+
```{ruby data_frame_no_a_column, warning=FALSE}
|
1759
|
+
df1 = R.data__frame(x: (1..3))
|
1760
|
+
puts df1
|
1761
|
+
```
|
1762
|
+
|
1763
|
+
Note that method mutate_y will fail independetly from the fact that variable 'a' is defined and
|
1764
|
+
in the scope of the method. Variable 'a' has no relationship with the symbol ':a' used in the
|
1765
|
+
definition of 'mutate\_y' above:
|
1766
|
+
|
1767
|
+
```{ruby call_mutate_y, warning = FALSE}
|
1768
|
+
a = 10
|
1769
|
+
mutate_y(df1)
|
1770
|
+
```
|
1771
|
+
## Different expressions
|
1772
|
+
|
1773
|
+
Let's move to the next problem as presented by Hardley where trying to write a function in R
|
1774
|
+
that will receive two argumens, the first a variable and the second an expression is not trivial.
|
1775
|
+
Bellow we create a data frame and we want to write a function that groups data by a variable and
|
1776
|
+
summarises it by an expression:
|
1777
|
+
|
1778
|
+
```{r diff_expr}
|
1779
|
+
set.seed(123)
|
1780
|
+
|
1781
|
+
df <- data.frame(
|
1782
|
+
g1 = c(1, 1, 2, 2, 2),
|
1783
|
+
g2 = c(1, 2, 1, 2, 1),
|
1784
|
+
a = sample(5),
|
1785
|
+
b = sample(5)
|
1786
|
+
)
|
1787
|
+
|
1788
|
+
as.data.frame(df)
|
1789
|
+
|
1790
|
+
d2 <- df %>%
|
1791
|
+
group_by(g1) %>%
|
1792
|
+
summarise(a = mean(a))
|
1793
|
+
|
1794
|
+
as.data.frame(d2)
|
1795
|
+
|
1796
|
+
d2 <- df %>%
|
1797
|
+
group_by(g2) %>%
|
1798
|
+
summarise(a = mean(a))
|
1799
|
+
|
1800
|
+
as.data.frame(d2)
|
1801
|
+
```
|
1802
|
+
|
1803
|
+
As shown by Hardley, one might expect this function to do the trick:
|
1804
|
+
|
1805
|
+
```{r diff_exp_fnc}
|
1806
|
+
my_summarise <- function(df, group_var) {
|
1807
|
+
df %>%
|
1808
|
+
group_by(group_var) %>%
|
1809
|
+
summarise(a = mean(a))
|
1810
|
+
}
|
1811
|
+
|
1812
|
+
# my_summarise(df, g1)
|
1813
|
+
#> Error: Column `group_var` is unknown
|
1814
|
+
```
|
1815
|
+
|
1816
|
+
In order to solve this problem, coding with dplyr requires the introduction of many new concepts
|
1817
|
+
and functions such as 'quo', 'quos', 'enquo', 'enquos', '!!' (bang bang), '!!!' (triple bang).
|
1818
|
+
Again, we'll leave to Hardley the explanation on how to use all those functions.
|
1819
|
+
|
1820
|
+
Now, let's try to implement the same function in galaaz. The next code block first prints the
|
1821
|
+
'df' data frame defined previously in R (to access an R variable from Galaaz, we use the tilda
|
1822
|
+
operator '~' applied to the R variable name as symbol, i.e., ':df'.
|
1823
|
+
|
1824
|
+
```{ruby r_dataframe}
|
1825
|
+
puts ~:df
|
1826
|
+
```
|
1827
|
+
|
1828
|
+
We then create the 'my_summarize' method and call it passing the R data frame and
|
1829
|
+
the group by variable ':g1':
|
1830
|
+
|
1831
|
+
```{ruby diff_exp_ruby_func}
|
1832
|
+
def my_summarize(df, group_var)
|
1833
|
+
df.group_by(group_var).
|
1834
|
+
summarize(a: :a.mean)
|
1835
|
+
end
|
1836
|
+
|
1837
|
+
puts my_summarize(:df, :g1)
|
1838
|
+
```
|
1839
|
+
|
1840
|
+
It works!!! Well, let's make sure this was not just some coincidence
|
1841
|
+
|
1842
|
+
```{ruby group_g2}
|
1843
|
+
puts my_summarize(:df, :g2)
|
1844
|
+
```
|
1845
|
+
|
1846
|
+
Great, everything is fine! No magic, no new functions, no complexities, just normal, standard Ruby
|
1847
|
+
code. If you've ever done NSE in R, this certainly feels much safer and easy to implement.
|
1848
|
+
|
1849
|
+
## Different input variables
|
1850
|
+
|
1851
|
+
In the previous section we've managed to get rid of all NSE formulation for a simple example, but
|
1852
|
+
does this remain true for more complex examples, or will the Galaaz way prove inpractical for
|
1853
|
+
more complex code?
|
1854
|
+
|
1855
|
+
In the next example Hardley proposes us to write a function that given an expression such as 'a'
|
1856
|
+
or 'a * b', calculates three summaries. What we want a function that does the same as these R
|
1857
|
+
statements:
|
1858
|
+
|
1859
|
+
```
|
1860
|
+
summarise(df, mean = mean(a), sum = sum(a), n = n())
|
1861
|
+
#> # A tibble: 1 x 3
|
1862
|
+
#> mean sum n
|
1863
|
+
#> <dbl> <int> <int>
|
1864
|
+
#> 1 3 15 5
|
1865
|
+
|
1866
|
+
summarise(df, mean = mean(a * b), sum = sum(a * b), n = n())
|
1867
|
+
#> # A tibble: 1 x 3
|
1868
|
+
#> mean sum n
|
1869
|
+
#> <dbl> <int> <int>
|
1870
|
+
#> 1 9 45 5
|
1871
|
+
```
|
1872
|
+
|
1873
|
+
Let's try it in galaaz:
|
1874
|
+
|
1875
|
+
```{ruby summarize_method}
|
1876
|
+
def my_summarise2(df, expr)
|
1877
|
+
df.summarize(
|
1878
|
+
mean: E.mean(expr),
|
1879
|
+
sum: E.sum(expr),
|
1880
|
+
n: E.n
|
1881
|
+
)
|
1882
|
+
end
|
1883
|
+
|
1884
|
+
puts my_summarise2((~:df), :a)
|
1885
|
+
puts "\n"
|
1886
|
+
puts my_summarise2((~:df), :a * :b)
|
1887
|
+
```
|
1888
|
+
|
1889
|
+
Once again, there is no need to use any special theory or functions. The only point to be
|
1890
|
+
careful about is the use of 'E' to build expressions from functions 'mean', 'sum' and 'n'.
|
1891
|
+
|
1892
|
+
## Different input and output variable
|
1893
|
+
|
1894
|
+
Now the next challenge presented by Hardley is to vary the name of the output variables based on
|
1895
|
+
the received expression. So, if the input expression is 'a', we want our data frame columns to
|
1896
|
+
be named 'mean\_a' and 'sum\_a'. Now, if the input expression is 'b', columns
|
1897
|
+
should be named 'mean\_b' and 'sum\_b'.
|
1898
|
+
|
1899
|
+
```
|
1900
|
+
mutate(df, mean_a = mean(a), sum_a = sum(a))
|
1901
|
+
#> # A tibble: 5 x 6
|
1902
|
+
#> g1 g2 a b mean_a sum_a
|
1903
|
+
#> <dbl> <dbl> <int> <int> <dbl> <int>
|
1904
|
+
#> 1 1 1 1 3 3 15
|
1905
|
+
#> 2 1 2 4 2 3 15
|
1906
|
+
#> 3 2 1 2 1 3 15
|
1907
|
+
#> 4 2 2 5 4 3 15
|
1908
|
+
#> # … with 1 more row
|
1909
|
+
|
1910
|
+
mutate(df, mean_b = mean(b), sum_b = sum(b))
|
1911
|
+
#> # A tibble: 5 x 6
|
1912
|
+
#> g1 g2 a b mean_b sum_b
|
1913
|
+
#> <dbl> <dbl> <int> <int> <dbl> <int>
|
1914
|
+
#> 1 1 1 1 3 3 15
|
1915
|
+
#> 2 1 2 4 2 3 15
|
1916
|
+
#> 3 2 1 2 1 3 15
|
1917
|
+
#> 4 2 2 5 4 3 15
|
1918
|
+
#> # … with 1 more row
|
1919
|
+
```
|
1920
|
+
In order to solve this problem in R, Hardley needs to introduce some more new functions and notations:
|
1921
|
+
'quo_name' and the ':=' operator from package 'rlang'
|
1922
|
+
|
1923
|
+
Here is our Ruby code:
|
1924
|
+
|
1925
|
+
```{ruby name_change}
|
1926
|
+
def my_mutate(df, expr)
|
1927
|
+
mean_name = "mean_#{expr.to_s}"
|
1928
|
+
sum_name = "sum_#{expr.to_s}"
|
1929
|
+
|
1930
|
+
df.mutate(mean_name => E.mean(expr),
|
1931
|
+
sum_name => E.sum(expr))
|
1932
|
+
end
|
1933
|
+
|
1934
|
+
puts my_mutate((~:df), :a)
|
1935
|
+
puts "\n"
|
1936
|
+
puts my_mutate((~:df), :b)
|
1937
|
+
```
|
1938
|
+
It really seems that "Non Standard Evaluation" is actually quite standard in Galaaz! But, you
|
1939
|
+
might have noticed a small change in the way the arguments to the mutate method were called.
|
1940
|
+
In a previous example we used df.summarise(mean: E.mean(:a), ...) where the column name was
|
1941
|
+
followed by a ':' colom. In this example, we have df.mutate(mean_name => E.mean(expr), ...)
|
1942
|
+
and variable mean\_name is not followed by ':' but by '=>'. This is standard Ruby notation.
|
1943
|
+
|
1944
|
+
[explain....]
|
1945
|
+
|
1946
|
+
## Capturing multiple variables
|
1947
|
+
|
1948
|
+
Moving on with new complexities, Hardley proposes us to solve the problem in which the
|
1949
|
+
summarise function will receive any number of grouping variables.
|
1950
|
+
|
1951
|
+
This again is quite standard Ruby. In order to receive an undefined number of paramenters
|
1952
|
+
the paramenter is preceded by '*':
|
1953
|
+
|
1954
|
+
```{ruby multiple_vars}
|
1955
|
+
def my_summarise3(df, *group_vars)
|
1956
|
+
df.group_by(*group_vars).
|
1957
|
+
summarise(a: E.mean(:a))
|
1958
|
+
end
|
1959
|
+
|
1960
|
+
puts my_summarise3((~:df), :g1, :g2)
|
1961
|
+
```
|
1962
|
+
|
1963
|
+
## Why does R require NSE and Galaaz does not?
|
1964
|
+
|
1965
|
+
NSE introduces a number of new concepts, such as 'quoting', 'quasiquotation', 'unquoting' and
|
1966
|
+
'unquote-splicing', while in Galaaz none of those concepts are needed. What gives?
|
1967
|
+
|
1968
|
+
R is an extremely flexible language and it has lazy evaluation of parameters. When in R a
|
1969
|
+
function is called as 'summarise(df, a = b)', the summarise function receives the litteral
|
1970
|
+
'a = b' parameter and can work with this as if it were a string. In R, it is not clear what
|
1971
|
+
a and b are, they can be expressions or they can be variables, it is up to the function to
|
1972
|
+
decide what 'a = b' means.
|
1973
|
+
|
1974
|
+
In Ruby, there is no lazy evaluation of parameters and 'a' is always a variable and so is 'b'.
|
1975
|
+
Variables assume their value as soon as they are used, so 'x = a' is immediately evaluate and
|
1976
|
+
variable 'x' will receive the value of variable 'a' as soon as the Ruby statement is executed.
|
1977
|
+
Ruby also provides the notion of a symbol; ':a' is a symbol and does not evaluate to anything.
|
1978
|
+
Galaaz uses Ruby symbols to build expressions that are not bound to anything: ':a.eq :b' is
|
1979
|
+
clearly an expression and has no relationship whatsoever with the statment 'a = b'. By using
|
1980
|
+
symbols, variables and expressions all the possible ambiguities that are found in R are
|
1981
|
+
eliminated in Galaaz.
|
1982
|
+
|
1983
|
+
The main problem that remains, is that in R, functions are not clearly documented as what type
|
1984
|
+
of input they are expecting, they might be expecting regular variables or they might be
|
1985
|
+
expecting expressions and the R function will know how to deal with an input of the form
|
1986
|
+
'a = b', now for the Ruby developer it might not be immediately clear if it should call the
|
1987
|
+
function passing the value 'true' if variable 'a' is equal to variable 'b' or if it should
|
1988
|
+
call the function passing the expression ':a.eq :b'.
|
1989
|
+
|
1990
|
+
|
1991
|
+
## Advanced dplyr features
|
1992
|
+
|
1993
|
+
In the blog: Programming with dplyr by using dplyr (https://www.r-bloggers.com/programming-with-dplyr-by-using-dplyr/) Iñaki Úcar shows surprise that some R users are trying to code in dplyr avoiding
|
1994
|
+
the use of NSE. For instance he says:
|
1995
|
+
|
1996
|
+
> Take the example of seplyr. It stands for standard evaluation dplyr, and enables us to
|
1997
|
+
> program over dplyr without having “to bring in (or study) any deep-theory or
|
1998
|
+
> heavy-weight tools such as rlang/tidyeval”.
|
1999
|
+
|
2000
|
+
For me, there isn't really any surprise that users are trying to avoid dplyr deep-theory. R
|
2001
|
+
users frequently are not programmers and learning to code is already hard business, on top
|
2002
|
+
of that, having to learn how to 'quote' or 'enquo' or 'quos' or 'enquos' is not necessarily
|
2003
|
+
a 'piece of cake'. So much so, that 'tidyeval' has some more advanced functions that instead
|
2004
|
+
of using quoted expressions, uses strings as arguments.
|
2005
|
+
|
2006
|
+
In the following examples, we show the use of functions 'group\_by\_at', 'summarise\_at' and
|
2007
|
+
'rename\_at' that receive strings as argument. The data frame used in 'starwars' that describes
|
2008
|
+
features of characters in the Starwars movies:
|
2009
|
+
|
2010
|
+
```{ruby starwars}
|
2011
|
+
puts (~:starwars).head
|
2012
|
+
```
|
2013
|
+
The grouped_mean function bellow will receive a grouping variable and calculate summaries for
|
2014
|
+
the value\_variables given:
|
2015
|
+
|
2016
|
+
```{r grouped_mean}
|
2017
|
+
grouped_mean <- function(data, grouping_variables, value_variables) {
|
2018
|
+
data %>%
|
2019
|
+
group_by_at(grouping_variables) %>%
|
2020
|
+
mutate(count = n()) %>%
|
2021
|
+
summarise_at(c(value_variables, "count"), mean, na.rm = TRUE) %>%
|
2022
|
+
rename_at(value_variables, funs(paste0("mean_", .)))
|
2023
|
+
}
|
2024
|
+
|
2025
|
+
gm = starwars %>%
|
2026
|
+
grouped_mean("eye_color", c("mass", "birth_year"))
|
2027
|
+
|
2028
|
+
as.data.frame(gm)
|
2029
|
+
```
|
2030
|
+
|
2031
|
+
The same code with Galaaz, becomes:
|
2032
|
+
|
2033
|
+
```{ruby advanced_starwars}
|
2034
|
+
def grouped_mean(data, grouping_variables, value_variables)
|
2035
|
+
data.
|
2036
|
+
group_by_at(grouping_variables).
|
2037
|
+
mutate(count: E.n).
|
2038
|
+
summarise_at(E.c(value_variables, "count"), ~:mean, na__rm: true).
|
2039
|
+
rename_at(value_variables, E.funs(E.paste0("mean_", value_variables)))
|
2040
|
+
end
|
2041
|
+
|
2042
|
+
puts grouped_mean((~:starwars), "eye_color", E.c("mass", "birth_year"))
|
2043
|
+
```
|
2044
|
+
|
2045
|
+
|
2046
|
+
[TO BE CONTINUED...]
|
2047
|
+
|
2048
|
+
|
2049
|
+
# Contributing
|
205
2050
|
|
206
2051
|
* Fork it
|
207
2052
|
* Create your feature branch (git checkout -b my-new-feature)
|
@@ -210,3 +2055,4 @@ puts gg
|
|
210
2055
|
* Push to the branch (git push origin my-new-feature)
|
211
2056
|
* Create new Pull Request
|
212
2057
|
|
2058
|
+
# References
|