Anonymous user
Talk:Welch's t-test: Difference between revisions
cannot use existing libraries
(cannot use existing libraries) |
|||
Line 484:
:A piece of general advice for your projects: don't reinvent the wheel, especially for numerical code. It's extremely difficult to get it right. And there are tons of free good libraries (start with Netlib, and the free matrix languages). And by the way, if the task is not too heavy, consider a matrix language in the first place, it's relatively slow, but it's extremely easy to work with. MATLAB is the gold standard, but the free ones are good too. If you have anything related to statistics, I strongly suggest R (and good news: linking C or Fortran code to R is fairly easy).
:[[User:Eoraptor|Eoraptor]] ([[User talk:Eoraptor|talk]]) 06:59, 14 December 2017 (UTC)
I always have difficulties with libraries, in a lot of ways they're like mythical creatures/unicorns to me. Consider http://people.sc.fsu.edu/~jburkardt/c_src/asa063/asa063.html which has C code (supposedly). None of it compiles! It's a lot like the user who overwrote my working Perl code to use some library I can't install. I couldn't get GSL to work either. What I wrote for the p-value calculation here may be awkward, but it at least does what it says it would do.
Consider: <code>con@e:~/Scripts/ASA$ gcc -o asa063 asa063.c -Wall -pedantic -std=c99 -lm
asa063.c: In function ‘betain’:
asa063.c:357:10: warning: unused variable ‘betain’ [-Wunused-variable]
double betain;
^
asa063.c: In function ‘timestamp’:
asa063.c:502:10: warning: variable ‘len’ set but not used [-Wunused-but-set-variable]
size_t len;
^
/usr/lib/gcc/x86_64-linux-gnu/5/../../../x86_64-linux-gnu/crt1.o: In function `_start':
(.text+0x20): undefined reference to `main'
collect2: error: ld returned 1 exit status</code>
I modified none of this. Perhaps you know how his asa063.c could compile?--[[User:Hailholyghost|Hailholyghost]] ([[User talk:Hailholyghost|talk]]) 13:28, 14 December 2017 (UTC)
| |||