@StatGenDan Thanks for linking this. It looks really interesting, and will form my "reward reading" when I'm done studying today.

(that's right - I reward study by doing more study lol)

I can't say whether or not I'll agree with the entire argument, but I can say that learning #baseR before introducing other packages suited me very well.

@edyhsgr Potentially controversial, but I agree!

I was annoyed when the code demos for a particular uni course relied on #tidyR, because I felt that I hadn't learned enough #baseR yet.

So I stuck to base for my big assignment, only bringing in other packages where needed. It's not that I dislike tidy; it's just that I want to be in control of the packages - I don't want the packages to be in control of me.

Get unique rows of a dataframe in #rstats, but ignore some columns (e.g., ID in this df)?

df1 <-
data.frame(
name = c("apple", "apple", "apple", "orange", "orange"),
ID = c(1, 2, 3, 4, 5),
is_fruit = c("yes", "yes", "yes", "yes", "yes")
)

If you're a #baseR fan like me, you might want:

df1[!duplicated(df1[!names(df1) %in% c("ID")]), ]

Major thanks to @tannguyen for spotting and reporting a bug in my niftiPlottingFunctions.R plot.volume() function!

The code at https://osf.io/k8u2c is updated; see https://mvpa.blogspot.com/2023/03/bugfix-for-niftiplottingfunctionsr.html for info and examples.

(This #baseR #rstats function is for plotting #fMRI #nifti images in #knitr documents.)

... it is painful to find out there was a bug, but so much better to fix it than to continue its propagation! I am grateful to @tannguyen and hope anyone else seeing something odd lets me know.

@milesmcbain "... the users are actually readers of the code rather than just runners ..." is a great way to put it!

With the #academic #programming that's my job, I assume most everything will eventually contribute to a publication, and so be shared. The logic of the code needs to be apparent to others in the future (and hopefully useable); imo avoiding dependencies and sticking to the most stable codebase (#baseR #rstats, in my case) is worth the effort.

A methods paper of mine is now at https://doi.org/10.3389/fnimg.2023.1070274 . This little #NewNeuroPaper demonstrates human task #fMRI quality control, as part of the FMRI Open QC Project (which I hope to discuss at #OHBM and elsewhere as it wraps up).

Its supplemental (https://doi.org/10.17605/OSF.IO/HT543) may be of interest to #rstats folks, particularly if looking for examples of #baseR #knitr #graphics and/or scripts aiming to minimize dependencies and maximize long-term usability.

@marcostaracido

Hai un ano estiven nunha situación kafkiana semellante, no meu caso coa compañía Unión #Fenosa Distribución e #Baser (unha comercializadora eléctrica de bono social) tirándose a pelota para non arranxar un problema de falta de potencia, isto no máis frío do inverno e pasando o COVID na casa.
A "atencion" telefónica, de tolos.
E o encanto de tecno-feudalismo que padecemos os súbditos dos #oligopolios .
¡Nacionalización de servizos básicos xa!

@ERDonnachie @tgtads @datamaps

I can't resist a #tidyverse vs. #baseR #rstats thread. 😅

It is totally possible to work without tidyverse, and base R graphics can be incredibly flexible. If you'd like some examples, just this morning I uploaded yet another set of all-base #knitr supplemental files (https://osf.io/ht543), with very pretty (and highly useful) graphics, if I say so myself. (Tidyverse can be ok, but it's not at all required.)

Maybe it’s time to remove dplyr from my Rprofile

https://lorentzen.ch/index.php/2023/01/09/dplyr-style-without-dplyr/

#Rstats #baseR #dplyr

@domliebl Nice #baseR graphing introduction!

I use base #rstats pretty much exclusively, and am probably more biased than you are against ggplot ... I used it for a few years when it first came out, but got frustrated by its limitations and the need for constant updating and dependencies.

If you are looking for base R graphics tutorials (plus #knitr), try https://mvpa.blogspot.com/2020/03/introductory-knitr-tutorial.html

Debate settled

What is better, base R of tidyverse?

#Rstats #baseR #tidyverse

As usually, the code for day 1 is still short enough to fit into a toot. And as usually, I try to solve the puzzles in #baseR until I get too lazy.

Day 1 was still a piece of cake in base R, with one of my all time favourites `by()` taking the center stage.

day1 <- data.frame(
calories = as.numeric(readLines("inputs/day01.txt"))
)

day1$elf_no <- cumsum(is.na(day1$calories))
day1 <- day1[!is.na(day1$calories), ]

calorie_sums <- by(day1$calories, day1$elf_no, sum)

max(calorie_sums)

#introduction

I'm a staff scientist at Washington U in St Louis (#Missouri, USA), mostly working with human task fMRI datasets. I'm a big fan of #baseR, #knitr, #QC, and very extensive #supplementary materials. I insist on looking at data early and often; some #tutorials and examples on my blog. (#GotBrains?)

I'm a #FirstGen college grad, #IowaStateU; BS in #forestry. I do #knitting (plus #crochet, #sewing, etc.), not just knitr-ing. I haven't before posted cat pictures, but no guarantees.

:rstats: `paste0()` can help you select columns in a df, base R style.

# generate example data
df <- as.data.frame(matrix(sample(30), nrow = 3))

# change last three column names
names(df)[ncol(df):(ncol(df) - 2)] <- c("some", "random", "names")

# switch columns up for fun :-)
df <- df[ , sample(10)]

# Extract columns V1-V7
df[paste0("V", 1:7)]

#rstats #baseR

#rstats #dataviz just gonna drop this here: I need to update this to the most recent data. Source was #MotherJones and is legitimate data. Not sure if this needs a content warning, but I'm not advocating for anything in particular. Just let the data speak for itself. If anyone wants the #Rcode I used, it uses #BaseR and I can put all of it into Github

Introduction:

Hi, I’m a political science postdoc researcher at the Computational Social Science Lab @Uni_Stuttgart

My roots are research on attitudes, #participation, and #voting behavior. My dissertation on participatory and #deliberative #democracy on local level brought me to python, webscraping and machine learning. #AI

I consider causal reasoning to be the most important part of statistical modeling. #DAG #DGP

What else to say? I love teaching #rstats the #baseR way!

ATM, we have approximately 6 people who use something other thand #dplyr, #datatable or #baseR. What do you use? #poorman?

#rstats