Timeline for Fast Hardy-Weinberg equilibrium simulation
Current License: CC BY-SA 4.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 2 at 15:18 | answer | added | Toby Speight | timeline score: 1 | |
| Aug 27, 2018 at 18:04 | comment | added | computergorl | @chux this version should be compilable. You would also need fast_rand.h which I linked to. I would recommend -msse4.2 -mpopcnt and -Ofast | |
| Aug 27, 2018 at 18:03 | history | edited | computergorl | CC BY-SA 4.0 |
added 4259 characters in body
|
| Aug 27, 2018 at 15:25 | comment | added | chux | "maximum number of organisms is 65536" appears to be off-by-one. Code looks like the description should be 65535? | |
| Aug 27, 2018 at 15:23 | comment | added | chux | To cursory and general comments about performance improvements does not rely on a post change. To make easy for reviewers and to provide good feedback does merit a change. Between copy-paste the whole thing and what is posted is a wide range of options that would improve. To make better, a reduced, yet compilable code snippet is work on someone's part. It adds clarity and efficiency if that is done by the poster. | |
| Aug 27, 2018 at 14:40 | comment | added | computergorl | @chux The reason I didn't define any object or function definitions is that I don't use any, other than the FastRand function which I explain at length and the popcnt instruction. The whole program is probably some 10 kilobytes to do everything, which I felt was too much to just dump on someone. Thus, I trimmed it to the section of code that runs over and over again and which needs optimization rather than the full program. Because that removes context, I tried to add the context back with the explanation. Would it have been better to simply copy-paste the whole thing? | |
| Aug 26, 2018 at 14:52 | comment | added | chux | @Vogel612 The post code lacks object and function definitions. From a C perspective, adding these are not burdensome for an OP - quite trivial in fact. My comment was not to suggest the code must compile and link, but to encourage OP to add enough, without undue overhead, to expiate good reviews - rather than none in 2 days. OP hints at wanting optimization. Cursory optimization is possible with incomplete code - as here. Testable and rateable optimization is possible when working code is posted here - thus making for a better question and review. | |
| Aug 26, 2018 at 10:10 | comment | added | Vogel612 | @chux there is no reqirement for the code to compile directly when copied from the question. That would place an unnecessary burden to, if not prohibit any question that contains a selection of classes (or possibly even a long single method) from a larger project. It's only required that the code be working to the best of the author's knowledge. | |
| Aug 25, 2018 at 23:50 | review | Close votes | |||
| Aug 26, 2018 at 10:10 | |||||
| Aug 25, 2018 at 23:32 | comment | added | chux | Posted code does not compile. Your question must contain code that is already working correctly | |
| Aug 24, 2018 at 15:00 | history | tweeted | twitter.com/StackCodeReview/status/1033006072179687424 | ||
| Aug 24, 2018 at 13:04 | comment | added | computergorl |
Sure, the complete program is here. There are no special libraries. To compile it and test for speed, use make speedtest, though you probably have to replace the gcc there with the gcc on your system. The full code is about 20kb so I didn't want to put up a ton of code that didn't really matter - I just wanted to focus on the core logic.
|
|
| Aug 24, 2018 at 10:43 | comment | added | Toby Speight |
You seem to have lost your #include lines and function declaration(s), perhaps while copying from your editor - any chance you could reinstate those, to give us a complete runnable program to reproduce your results?
|
|
| Aug 24, 2018 at 4:50 | history | edited | 200_success | CC BY-SA 4.0 |
added 11 characters in body; edited tags
|
| Aug 24, 2018 at 4:37 | history | asked | computergorl | CC BY-SA 4.0 |