rmhogervorst / cleancode Goto Github PK

Merging van verschillende soorten csv's in verschillende formaten. oude coffee data, dan nieuwe vanuit taplog.

Merging van verschillende soorten csv's in verschillende formaten. oude coffee data, dan nieuwe vanuit taplog.

reading in data, meta thing from here: http://www.r-bloggers.com/importing-data-into-r-part-two/

rio package

readxl
readr

All sorts of examples. With numbers, letter @ dots $ quotes many matches specific

short walkthrough. collected works in templates,
gitignore. readme.

some cool tricks, .github folder

F.i. :
Did you like it? Do you have some idea's? post them in the issue tracker here:

using sourcetree for changes.

introduction to what all the tools are, as replacement for word reference managers.

set up: collect and annotate documents in zotero, save to bibtex file.

write in rmarkdown and add citations with button, can search in it
was already possible to use citations, but is a bit annoying.
http://rmarkdown.rstudio.com/authoring_bibliographies_and_citations.html

therefor button is really nice.

links to other projects:
https://github.com/scholmd/scholmd

just for fun

keeps everything together.

beginner

expand with veracrypt for protection of sensitive data

Had nog een paar dingetjes die bij mij niet werkten, mocht je die nog willen aanpassen voor een niet-beta versie:

•      In sommige van de voorbeeldscriptjes staat nog een ‘r’ voordat het scriptje begint. Op het moment dat deelnemers die mee kopiëren werkt het script niet
  

•      sapply(wo2015, class) head(wo2015) staat nu achter elkaar, maar moeten op een andere regel omdat het om twee verschillende commands gaat -> misschien dat in de tekst duidelijker maken?
  

•      Attributes (wo2015$OPLEIDINGSVORM) geeft mij als uitkomst ‘NULL’. Attributes(wo2015[1,2]) werkt wel.
  

•      Wanneer ik wo2015$OPLEIDINGSFASE.ACTUEEL <- as_factor(wo2015$OPLEIDINGSFASE.ACTUEEL) intyp krijg ik als error: ‘Error in sort.list(y): invalid input 'initiële opleiding' in 'utf8towcs'’
  

•      Ik krijg steeds foutmeldingen als ik opleidingsfase.actueel2 wil gebruiken in een van de scripts, en als ik in het SPSS bestand kijk zit die er ook helemaal niet tussen, kan het zo zijn dat je gewoon actueel bedoelde ipv actueel2?
  

o Als ik ‘actueel’ gebruik lukt het me wel om het laatste script te runnen, maar het groupen op meerdere levels met meerdere arguments (een-na-laatste opdracht) lukt nog steeds niet.

• rstudio tips ('####' ) and other double things section
• New things in Rstudio tearing up and stuff

Packrat
docker / rocker

automated builds with travis.ci

don't take on too many dependencies (import them into your project)

use datasets already used and transform to chart.

f.i. schooldata of universities over time
data from secondary schools in one province, south holland and map to approx locations.

http://blog.jupo.org/2013/05/05/sandboxed-jekyll-hacks/ for instance:

<ul>
{% for post in site.posts %}
    {% assign post_words = 0 %}
    {% assign lines = post.content|split:'pre>' %}
    {% for line in lines %}
        {% assign mod = forloop.index|modulo:2 %}
        {% assign line_words = line|strip_html|number_of_words|times:mod %}
        {% assign post_words = post_words|plus:line_words %}
    {% endfor %}
    <li><a href="{{ post.url }}">{{ post.title }}</a>({{ post_words }} words)</li>
{% endfor %}
</ul>

Change this blog, idea's see something weird? change it on github

at bottom of page? or top, link to page

devtools:: session_info() locale, active packages, versions. R version , windows 10 64 bit.

Other relative links.

THIS IS A TEMPLATE, IF YOU DON'T HAVE AN ISSUE WITH THE WEBSITE
OR SOMETHING ELSE YOU CAN DELETE EVERYTHING AND TYPE WHAT YOU WANT.
I MEAN, YOU CAN ALWAYS DO WHAT YOU WANT, I'M A TEMPLATE, NOT A POLICEMAN.
Hi Roel,

I saw your post on rmhogervorst.nl/cleancode / r-bloggers,

THIS IS WHERE YOU WRITE DOWN WHERE I SCREWED UP, MADE A TYPO OR WHAT EVER.

IF YOU HAD AN ERROR FOLLOWING A TUTORIAL

I tried to [what you wanted to do]
by doing the followings steps
[steps]

But it didn't work/ it failed miserably/ It broke my computer and my life's work/ etc
This happened:

[error messages, results etc]

I have used version
[in rstudio you can find the version of the package in the packages pane/ or use sessionInfo() for all active packages
or you can use packageVersion("packagename") for the specific version of the package]
with R version [ version]
on a windows / mac / linux / freebsd / solaris, etc computer.

scripts, scripts in 1 keer uitvoeren als functie -> opdelen in logische partities -> je functies algemener maken - functies in package

H(o)i Roel,

I saw your nice post about maps on rbloggers. I especially liked the way you introduced SpatialPolygonsDataFrames.

Just wanted to let you know I wrote a package for visualizing maps, called tmap (https://github.com/mtennekes/tmap), that turned out to be pretty ambitious. It works like gglot2, but overcomes some of drawbacks. Here's a little example:

library(tmap)
data(NLD_muni, data(NLD_prov)

tm_shape(NLD_muni) +
  tm_polygons("population", convert2density=TRUE) +
tm_shape(NLD_prov) +
  tm_borders(lwd=2)

# interactive mode
tmap_mode("view")
last_map()

If you have any questions or suggestions, please let me know.

Best,
Martijn

git on networkdrive

http://www.watkyn.com/blog/2011/09/22/Using-git-at-work-on-a-Windows-network-drive/

Hi Roel,

I like your post on rmhogervorst.nl/cleancode: http://rmhogervorst.nl/cleancode/blog/2016/06/13/NSE_standard_evaluation_dplyr.html .

I have a package replyr up on CRAN that helps with this task and a free video demonstrating how to use the package https://youtu.be/iKLGxzzm9Hk . I'd love for you to give replyr::let a look and perhaps recommend it.

So its more like real world data analysis work, 80% cleaning.

see by plotting

with gapminder data over time?
with studies over time numbers in the netherlands per university

eerst deze:

http://kbroman.org/pkg_primer/pages/tests.html
dan http://r-pkgs.had.co.nz/tests.html

misschien apart testonderdeeltje maken?

voorbeeld: dat output ineens anders was, assertr
past in grotere gedeelte van hoe zorgen we dat code future proof blijft

Tips om over na te denken, voor schrijven blogpost, er zijn 3 verschillende manieren hoe mensen het web gebruiken voor programmeren, uitleg (tutorials, just in time learning), clarification (begrijpen het het werkt in een andere taal) en reminder (als hulpstukje om naar terug te gaan voor gebruik). Andere dingen: veel tekst is niet handig, slaan ze vaak over. willen graag uitproberen. ref: "Two studies of opportunistic programming: interleaving web foraging, learning, and writing code " in zotero onder programming

• add to .Rprofile with this source

https://github.com/yihui/fun

http://jtleek.com/protocols/travis_bioc_devel/ veel bredere uitleg van hoe je met travis moet werken.

hoe zet je het op en wat komt eruit?

Create post updated nse with dplyr for version 6.
Intro: Link back to old post. And to the suggestions.

Main part: per verb
Select, filter, rename, group by, summarize.
Old way NSE, other package, new style NSE

Conclusions: new style programming is probably better.

[ ] explain what rhistory file is
[ ] idea of sensitivity of that file (shiny idea)
[ ] find all your .rhistory files in your projects
[ ] load all of the files in list
[ ] search for common patterns
[ ] find common things / errors

tests, t-test, anova, regression, normality tests.

https://www.google.nl/url?sa=t&rct=j&q=&esrc=s&source=web&cd=6&ved=0ahUKEwjc1fextcjOAhXDthoKHcqaDsMQFggtMAU&url=http%3A%2F%2Fextremepresentation.typepad.com%2Ffiles%2Fchoosing-a-good-chart-09.pdf&usg=AFQjCNFzYPglkr0MXK9x0r94WosNgjAEPA&sig2=X2sBsu43re_mkmITNKyM-Q&cad=rja

extremepresentation.typepad.com/.../choosing-a-good-chart-09.pd...

Nice post! But the figures do not display either on my mobile, IE,or in chrome.

Shouldn't it be

Tips about clean R-code and reproducible work

?

show distinct population and 2 seperate but close to eachother. see effect of t-test mean.

http://www.r-bloggers.com/deploying-desktop-apps-with-r/ standalone desktopapps

I've been trying to use shapefiles and plot maps for quite a while. This is the first time I really rolled my sleeves up. Your codes worked smoothly. Thank you for sharing.

I have a question about some of the details you used in your workflow. Hope you could give some advices.

Excluding inland waters
You removed lakes by excluding their names, which is convenient only if one knows and the data provide the lake names. Besides, there are also rivers (canals in NL?) in land. Fortunately there are shapefiles for waters that I can find online, which, like the land maps, provide polygon info for inland waters. If I cover the waters polygons in front of land polygons, with different color, it looks nice. Thing is, the waters polygons do not include administrative information, which is understandable since a river can cross many places. But when I try to plot a map for the land and waters of an individual province, I can not exclude waters from other provinces. This makes the whole picture much larger than the specific province which I intend to plot. Do you have any suggestion on this?

Thank you.

• gitignore (https://www.gitignore.io)

there are a gazillion really great courses( see below 1, 2, 3, 4). link to own courses...

The goal of these three lessons <- there are 4

very useful shell around R <- could change the wording