People think R is just for statistics, but its capabilities go far beyond that. The data wrangling, large-scale number crunching, visualization ecosystem, and especially the Shiny library for building interactive web applications make it incredibly powerfulI recently built a complex Shiny app with help from LLMs within a day and demoed it to my team. It reduced the workflow time by more than 90%. Now other departments are on board, and even managers want to understand what tool I used.Thoughts?
shiny is a library and not a language feature. Shiny is pretty stupid, why not keep a minimal CLI? Sounds like something stupid non-technical people would praise
>>108784104>>Thoughts?normies hate hard work >even managers want to understand what tool I used.don't tell them
>>108784214why is shiny stupid? you just say without giving any justification.
>>1087842311. it created massive overhead, you need a server, render html files, etc2. a simple script file with command line arguments can be intergated easily with other programs and used in automated workflows3. when you say it reduced workflow time by 90%, the remaining 10% are likely the UI clicks. Are you too stupid to fully automate the task?
>>108784104Everyone I know who uses R a lot, also uses python a lot. Sometimes other languages not to mention libraries written in C and Fortran.In general, R suffers just like every other niche language against a mega-popular one. But it has a lot of nice traits. Perhaps the nicest thing about R compared to python is the long-term stability. Python loves removing standard libraries and changing apis and adding superfluous shit like color repl and error messages while R is more committed to not jerking around the researchers who use it.
> built a shiny app> no screenshots providedkeep on the cosplaying
>>108785367>Python loves removing standard libraries and changing apis and adding superfluous shit like color repl and error messagesI've been using python daily for like what, 11 years now? And I barely notice any of this crap people are complaining about. Literally just>use a few stable, well-maintained libraries>write your own fucking implementation code rather than leftpadding everything>use a normal operating systempython is slow but it's a proper programming language that interfaces with your system very naturally. it is also easy to integrate with C++ and C packages. R is an academic tool. I think most of the retards who struggle with python are overly fixated on enterprise software development concepts like DRY. autistically trying to encapsulate every little bit of implementation code rather than recognizing that you can often roll a local version of whatever you need faster than figuring out the right dependencies to use, because of how expressive the language is
>>108785705I've been writing python since 2006, even had programs that were 2+3 compatible for awhile. I know how to avoif dependencies but R is still more stable, especially when you include the ecosystem. And unfortunately a lot of work does require using 3rd party libraries and api wrappers.
