The R Project SVN R

Rev

Show changed files | Details | Compare with Previous | Blame | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
47757 6391 d 13 h jmc /branches/R-classInheritance/ create branch for class inheritance  
46101 6580 d 21 h maechler /trunk/src/nmath/ (long-standing) cosmetic  
44785 6707 d 18 h ripley /trunk/ clean up distribution API to match R one where easy to do.  
43893 6777 d 20 h maechler /trunk/ pnchisq(): 2nd part of fixing PR#10551  
42687 6908 d 6 h ripley /trunk/src/nmath/ use long double accumulators where available  
40105 7178 d 16 h murdoch /trunk/ pnchisq gave the wrong tail in extreme cases (PR#9406)  
37461 7454 d 0 h ripley /trunk/src/nmath/ report where ML_ERROR is occurring  
37290 7476 d 5 h ripley /trunk/src/nmath/ copyright statements  
37287 7476 d 14 h ripley /trunk/ improvements to [pq]nchisq
- use direct formula which allows for lower_tail = FALSE if ncp < 80
(this is often a lot more accurate).
- use starting point and lower_tail in qnchisq
- can be slower, so make interruptible		  
37281 7477 d 2 h ripley /trunk/ Get ML_ERROR to do something useful, and ensure MATHLIB messages end in \n.
Warn on cancellation errors in non-central CDFs for t, f, beta, chisq		  
37223 7483 d 1 h ripley /trunk/src/nmath/ improve documentation  
36986 7509 d 23 h ripley /trunk/ port.c was using d1mach and i1mach even though they were declared private
more use of visibility attributes		  
32876 7853 d 0 h ripley /trunk/ mark messages  
28809 8162 d 19 h maechler /trunk/ major pnchisq() {and some qnchisq()} improvements for larger ncp  
24049 8496 d 17 h ripley /trunk/ branch update  
13115 9269 d 3 h maechler /trunk/src/nmath/ qchisq(*, ncp, lower) fix {+ improvement else}  
11412 9385 d 1 h hornik /trunk/src/ Use `chi-squared'.  
10470 9463 d 20 h maechler /trunk/src/ Rboolean & TRUE & FALSE  
9363 9556 d 19 h ripley /trunk/ standalone Mathlib  
8431 9643 d 19 h ripley /trunk/ use a private header file for Mathlib, several small consistency fixes  

←Prev12Next→

Show All