1.11 make Array a native julia type instead of a c wrapper, I think Dictonnary now also rely on the new Memory type so that made it even better. 1.12 should allow the compiler to actually use it to say if a vector should be only stack allocated ect ect so maybe another 1.5 x there ? not sure at all
I think you’re referring to the table function in base R. Yeah, you could load up all words in a vector and then pass that vector into the table function and then use the names function to get the words out of the table result (as the table result itself holds the numbers).
1.11 make Array a native julia type instead of a c wrapper, I think Dictonnary now also rely on the new Memory type so that made it even better. 1.12 should allow the compiler to actually use it to say if a vector should be only stack allocated ect ect so maybe another 1.5 x there ? not sure at all
Sounds cool!
What would the same task in R look like given the native aggregation functions ?
I think you’re referring to the table function in base R. Yeah, you could load up all words in a vector and then pass that vector into the table function and then use the names function to get the words out of the table result (as the table result itself holds the numbers).
@@ekbphd3200And speed-wise?
I haven’t tested it, but I assume it would be slower than data.table and tidyverse.
That is not even an order of magnitude difference… and the Python code is not even optimized for speed. Well done Python! 🎉
Very true!