R^4 #15: tidyverse and data.table, sitting side-by-side ... (part 1) Using a simple example to compare and contrast two approaches goo.gl/o92xJe #rstats h/t @gelliottmorris

Jan 21, 2018 路 11:44 PM UTC

9
26
4
64
Load newest
Replying to @eddelbuettel @gelliottmorris
Results of the profmem package: total(profmem(transformTV(rdin))) - 1475576 total(profmem(transformDT(dtin))) - 1241504
2
Replying to @gelliottmorris @hrbrmstr @bhive01 @eddelbuettel
Every time verse undoubtedly loose people start looking for new excuse. Myth about syntax is most boring. Common guys you can do better.
4
The pace of change is the main reason I don't use anything from the 'verse. I don't have time to fix my code because something has changed to make it easier for beginners.
11
Introverts unite!
Replying to @eddelbuettel @gelliottmorris
The point about needing more code is misleading - the tidyverse code is clearly from a notebook format, but for fair comparison should be expressed as one continuous pipe. I edited both down to 16 lines.
1
1
3
Replying to @eddelbuettel @gelliottmorris
I've never understood TV complaints about DT syntax -- I immediately took to DT syntax and still have yet to come across something I found so natural/intuitive in any language. Pipes, on the other hand, took me quite some time to naturalize 馃槒
1
3
Replying to @eddelbuettel @gelliottmorris
Nice! (except the nonsense comment about religion). DT is generally faster and has a fascinating syntax.
1
1
3