I just linked that because I wrote it and it came out yesterday. 
-
-
Replying to @o_guest
I'm reading it now. I find the point about not needing original code interesting. I've heard people say w/r/t analyses that >
1 reply 0 retweets 1 like -
> reproducibility is all about being able to take someone's code and get the same result. I prefer your definition, though.
1 reply 0 retweets 1 like -
Replying to @ed_berry
well that definition is out there just not applicable (imho) to computational modelling since for us the model is the experiment
1 reply 0 retweets 1 like -
Replying to @o_guest
True. For experimental stuff the data and the description of the analysis in the paper should be sufficient, imo.
3 replies 0 retweets 0 likes -
There is definitely too much incomplete description of analyses going round
1 reply 0 retweets 0 likes -
Replying to @ed_berry
again imho I don't think that's the crux but it's important of course. If an effect exists it should exist regardless of analysis
1 reply 0 retweets 1 like -
super tired, will crash, but I've a paper (2nd author) that covers my views regarding models and why they need to be rewritten
2 replies 0 retweets 0 likes -
no pressure for you to look at it just much easier than tweets late at night when I should be unwinding haha
1 reply 0 retweets 0 likes -
catch you later on!
super nice charts!
night!
2 replies 0 retweets 0 likes
*chats lol
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.