R Function A Day

As requested by some of you, there is now a book of these posts! 📗 bookdown.org/IndrajeetPatil… Makes it easy to- 📑 read 🔍 search 🔗 share etc. It ain't pretty, but that's the best I'd do in a day 😅 PRs welcome if you notice that something is amiss 🙏 #rstats #DataScience

This function no longer resides in {parameters} 📦. It can now be found in {datawizard} 📦: easystats.github.io/datawizard/ref…

If you want to get a quick overview of descriptive statistics for a numeric variable, {describe_distribution} from {parameters} 📦 is your friend 🙌 easystats.github.io/parameters/ref… Also works with a {grouped_} dataframe from {dplyr} 👏 #rstats #DataScience

This function no longer resides in {parameters} 📦. It can now be found in {performance} 📦: easystats.github.io/performance/re…

Factor analysis (FA) can help reduce many features to a few latent features. But one first needs to check if data is suitable for FA. The {check_factorstructure} function from {parameters} provides an informative and verbose way 🔍 easystats.github.io/parameters/ref… #rstats #DataScience

@IsmaelLCalandri @patilindrajeets

@rfunctionaday @patilindrajeets So sad. Your tweets were the first good thing of the day every day.

Today concludes this 1-year hobby project. There won’t be new posts, so feel free to unfollow, but do keep revisiting! 🙃 With > 22K followers in 365 posts, I guess you enjoyed reading them as I much as I’d preparing them. 😅 Best luck with your jouRneys! 🤞 @patilindrajeets

@malte_grosser @patilindrajeets Ask and ye shall receive 😅 bookdown.org/IndrajeetPatil… Thanks a lot for the suggestion! 🙌

@rfunctionaday @patilindrajeets Really cool project. Why not finish this by converting it into a bookdown?

@malte_grosser @patilindrajeets Good idea! Might work on that soon, but until then this timeline widget should come in handy 🧵 publish.twitter.com/?query=https%3…

@Bishal_Bhardwaj @patilindrajeets The tweets from the past will always be accessible. They will neither be removed nor changed in any way.

@rfunctionaday @patilindrajeets Where can we access these post when we need it? Yes we enjoyed and wait for another hobby project that helps we like learner.

@GentForAnth @patilindrajeets

@fedenichetti @patilindrajeets

@rfunctionaday @patilindrajeets Worst news since the pandemic started.

If we need to bind multiple dataframes by rows, we may first wish to check if this can be done successfully. The {compare_df_cols} function from {janitor} 📦 provides a summary of data types to check this! 🍎🍏 sfirke.github.io/janitor/refere… #rstats #DataScience

@JonGretar @kiwiSteveTaylor @rstats4ds In real use cases, this routine will actually be used to access a variable - specified by the user - from a data frame. Note that now `data$x` won’t work because we don’t know a priori what `x` is going to be.

@kiwiSteveTaylor @rfunctionaday @rstats4ds I guess a real world use case would involve a longer pipe.

While working in a piped workflow, if you need to extract a single column, the syntax to do so can be a bit cumbersome. The {pull} function from {dplyr} 📦 provides a more readable syntax to this end! 💄 dplyr.tidyverse.org/reference/pull… #rstats #DataScience

@ThomVolker Yes, I have shared it in the past 👇 But, AFAIK, that operator is used for side effects. So you will be able to print that column but won’t be able to assign it to an object. I think.

Visualizing data along an Archimedean spiral can efficiently reveal periodic patterns in time series data. The {spiral_*} function family from {spiralize} 📦 draws such spiral plots flexibly! 🌀 jokergoo.github.io/spiralize_vign… #rstats #DataScience

Serialization changes objects to a byte stream that can be saved to a binary file, while deserializing does the reverse. The {q*} function family from {qs} 📦 provides a performant way to serialize or deserialize any R object! ⚡️ rdrr.io/cran/qs/man/ #rstats #DataScience

Sometimes you may wish to include some patterns or images in plots for enhanced storytelling. The {image*} or {pattern*} function family from {patternplot} 📦 provides a way to do so! ⚜️ rdrr.io/cran/patternpl… #rstats #DataScience

@Fischuu None at all. In fact, AFAIK, the implementation of this data type in {reticulate} is based on names lists. Dictionary just provides a familiar name for bilingual users to a data structure that is not native to R but can be easily implemented.

@rfunctionaday What is the advantage of py_dict over named lists from base-R?

Python dictionary is an unordered data type with key-value pairs that allows accessing values, not by indexing, but via unique keys. The {py_dict} function from {reticulate} 📦 gives you access to this data type in R! 📖 rstudio.github.io/reticulate/ref… #rstats #DataScience

Sometimes, especially in the context of statistical modeling, there might be infinite or NaN's present in outputs that we may wish to replace with NAs. The {zap_inf} helper function from {sjmisc} 📦 does exactly this! ♾ strengejacke.github.io/sjmisc/referen… #rstats #DataScience