Jakob Richter

Machine Learning, Statistics, Photography

[Jakob Richter @jak0br]

Join us at the #Berlin R Users Meet-Up to talk about #MachineLearning in R. Sebastian Fischer and me will present the current state of ML Pipelines in #mlr3.
📅Tue, 21th Nov 6PM 📍@spicedacademy #RStats #meetup #DataScientists

[Jakob Richter @jak0br]

Amazing to see your papers being published after being out of academia for a while. A big thank you to all collaborators!

[Jakob Richter @jak0br]

Is there a dev workflow where you write software specifications under version control and a PR to this specification is linked to a PR in the code? It feels wrong to have specs only in tickets/tasks.

[Jakob Richter @jak0br]

@BR_AILab You are paying for rstudio pm? Installing the packages to a directory that is a persistent mount should already avoid unneeded recompilations.

[Jakob Richter @jak0br]

@antoine_fabri But SO is also pretty democratic. Even I have the right to decide about closing questions. Maybe I should use this privilege more often to help the moderators.

[Jakob Richter @jak0br]

@PhDemetri Linear models are for inference on the functional relationship between features and label. LMs can give you insights that a xgboost model couldn’t. And know you also know that the relationship is not that linear as you assumed.

[Jakob Richter @jak0br]

@FabioDeMasi Stimmt, aber in dem Ausschnitt geht unter anderem auch darum, dass er offensichtlich bei einer eidesstattlichen Versicherung gelogen hat.

[Jakob Richter @jak0br]

@eddelbuettel @patilindrajeets I think it's good practice to only use isTRUE/isFALSE if you expect a non TRUE/FALSE value. In complex scenarios, you are likely to suppress important warnings/errors that can make you aware of bugs in your code.

[Jakob Richter @jak0br]

@danluu I learned the hard way that there is a difference between realizing the obvious and remembering the obvious. Since then, I am thankful for each well phrased repetition and "make aware"-ings. People should appreciate the work of summing-up, otherwise all knowledge gets lost.

[Jakob Richter @jak0br]

@ScienceThekla @eddelbuettel Keine Ahnung, um welche Frage es geht. Vielleicht einfach mal fragen und schauen, ob jemand antwortet?

[Jakob Richter @jak0br]

@ScienceThekla @eddelbuettel Wenn es deutsch sein muss, dann ist das https://t.co/C8Hiv18B3p vielleicht eine gute Anlaufstelle? Oder https://t.co/FjvcgkHMzu ? Fühlt sich beides sehr 00er an.

[Jakob Richter @jak0br]

A colleague created this very nice and sleek weighted voting website for consensus decision-making. You might want to try it for your next group poll on where to go for dinner, or even for something more serious. https://t.co/IZ9PW3y91Q #konsensit

[Jakob Richter @jak0br]

@simonw Communication.

[Jakob Richter @jak0br]

@fchollet The problem arises if you don’t have a mental model yet.

[Jakob Richter @jak0br]

@antoine_fabri @TimTeaFan @rstats4ds Nice, thanks for using dplyr_reconstruct as well😀

[Jakob Richter @jak0br]

Often I want to merge two tables and A) make sure that all left entries are used when using an inner join or B) make sure that no right entries are left out when using a left join. Is there a function for it? Would there be an interest in a tiny R-Package? #RStats #DataScience

[Jakob Richter @jak0br]

We often have these assumptions that the left side is a superset of the right side. Detecting violations of such assumptions might save you a lot of trouble, when processing data sets.

[Jakob Richter @jak0br]

“Public key” is kind of an oxymoron to most peoples perception of a key. Naming it “private key” and “public lock” would probably have helped. 🤔

[Jakob Richter @jak0br]

@HeidiBaya Publishing to closed access is done for status and not publishing a slightly altered version later is just laziness or due to scare tactics of publishers. Universities should educate their researchers about their rights.

[Lukas Pfahler ([email protected]) @whadupapp]

"There’s no scientific integrity if you can’t control for human desperation." https://t.co/6DbaIBF6Z8