tag:blogger.com,1999:blog-4113082053753185517.post7993434466088984544..comments2023-07-02T13:50:28.495+02:00Comments on Samuel Bosch: Benchmarking reading binary values from a file with F#, Python, Julia, R, Go and OCamlSamuel Boschhttp://www.blogger.com/profile/00902576471871535529noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-4113082053753185517.post-92019405581652859562015-12-01T13:21:04.826+01:002015-12-01T13:21:04.826+01:00All my tests run on the same set of abaout 40 file...All my tests run on the same set of abaout 40 files but the Julia code from Ismael VC was only benchmarked on one file so for example when reading 10,000 times 10 values a total of 40 * 10000 * 10 = 4.000.000 numbers are read in my tests. All my code can be found here https://github.com/samuelbosch/blogbits/blob/master/geosrc/binreaders/Samuel Boschhttps://www.blogger.com/profile/00902576471871535529noreply@blogger.comtag:blogger.com,1999:blog-4113082053753185517.post-37966846422308315022015-12-01T13:11:47.342+01:002015-12-01T13:11:47.342+01:00Sorry not clear! So you read 40 files in Jullia an...Sorry not clear! So you read 40 files in Jullia and R but only one file with others. Is that what leads to the time difference or unfair comparison? ThanksRajanhttps://www.blogger.com/profile/13982834240487536694noreply@blogger.comtag:blogger.com,1999:blog-4113082053753185517.post-21078031865589821962015-11-24T11:03:19.330+01:002015-11-24T11:03:19.330+01:00Thanks for the effort. I've tested your code o...Thanks for the effort. I've tested your code on my benchmark and the performance is the same as my <a href="https://github.com/samuelbosch/blogbits/blob/master/geosrc/binreaders/binreader.jl" rel="nofollow">original version</a>. Initially I used Julia 0.3 and I've tested it on Julia 0.4.1 but that doesn't seem to change anything. Note that in my tests I read multiple files (nearly 40), which corresponds with the ~40 fold difference between your and my measurements.Samuel Boschhttps://www.blogger.com/profile/00902576471871535529noreply@blogger.comtag:blogger.com,1999:blog-4113082053753185517.post-72685612557368371602015-11-21T04:33:30.349+01:002015-11-21T04:33:30.349+01:00Hi I just saw your benchmarks and noticed the bad ...Hi I just saw your benchmarks and noticed the bad timings for Julia, there isn't a date of when you published this, but I tried to improve the Julia implementation: http://git.io/v4xgA<br /><br />10,000 times 10 values:<br />0.794152 seconds (487.72 k allocations: 26.247 MB, 0.24% gc time)<br /><br />10 times 10,000 values:<br />1.480457 seconds (667.21 k allocations: 34.061 MB, 0.43% gc time)<br /><br />Using a 4 cores Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz, Windows 7, Julia 4.0<br /><br />Which makes more sense!<br />Anonymoushttps://www.blogger.com/profile/17808566072090099984noreply@blogger.com