Hi!
How can I get a P value for a Spearman's rho (like in R's cor.test) in Julia? I'm using StatsBase's corspearman and RCall isn't an option since I'll run the test in parallel. Best, -- You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Unfortunately, I don't think this has been implemented yet so either you'd have to save all the correlations and use RCall afterwards or you'd have to implement the test. On Thu, May 19, 2016 at 8:25 AM, Diego Javier Zea <[hidden email]> wrote: Hi! You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Can I write a Julia version for HypothesisTests or StatsBase based on /R-3.3.0/src/library/stats/src/prho.c ? The license of the R's file is:
I was trying to do my own version following: Vettering; Teukolsky; Flannery (1992). Numerical Recipes in C: The Art of Scientific Computing (2nd ed.). p. 640. But I don't get the same values than R (I get the same number in some situation). My code is:
Thanks! You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
No. Unfortunately, neither of the sources can be used for an implementation here because of their licenses. You could try to write an implementation based on "The Upper Tail Probabilities of Spearman's rho" without using the actual printed implementation in the paper. On Thu, May 19, 2016 at 1:58 PM, Diego Javier Zea <[hidden email]> wrote:
You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Thanks!
-- "For n > 7 a power series expansion in n-1 based on the Edgeworth series approximation (David et al., 1951) is used." I'm reading the David's papers, but I don't understand exactly how to perform the approximation... I'm going for the R/RCall solution right now. Best, You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
In reply to this post by Andreas Noack
Can you elaborate on why you think an implementation based on the printed formulas in a paper could not be used? That does not sound correct to me (I'm a scientific journal editor). R itself relies heavily on algorithms from published papers, properly cited in the documentation for each function. On Thu, May 19, 2016 at 8:04 PM, Andreas Noack <[hidden email]> wrote:
You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
An implementation based on formulas in a paper is fine, but an implementation derived from GPL code must be GPL. R is GPL so has no issue with using GPL code. On Thu, May 19, 2016 at 4:26 PM, Michael Krabbe Borregaard <[hidden email]> wrote:
You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
In reply to this post by Michael Borregaard
The actual Fortran implementation of the algorithm is printed in the paper. That is the part I was talking about. It's not a problem to write an implementation based on the formulas. On Thu, May 19, 2016 at 4:26 PM, Michael Krabbe Borregaard <[hidden email]> wrote:
You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Cool, thanks for clearing that up. I am home without my uni VPN so could only read the first page of the paper... On Thu, May 19, 2016 at 10:33 PM, Andreas Noack <[hidden email]> wrote:
You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Will be OK if I follow the Numerical Recipes' formulas and explanations but not its code? Thanks! 2016-05-19 17:43 GMT-03:00 Michael Krabbe Borregaard <[hidden email]>:
Diego Javier Zea Departamento de Ciencia y Tecnología Universidad Nacional de Quilmes Roque Saenz Pena 182 1876 Bernal Argentina Phone:+541143657100 ext 4135 You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
In reply to this post by Michael Borregaard
I sent my implementation of Spearman test to HypothesisTests. It's based on the formulas (but not in the code) of Numerical Recipes and other references: Best, You received this message because you are subscribed to the Google Groups "julia-stats" group. To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email]. For more options, visit https://groups.google.com/d/optout. |
Free forum by Nabble | Edit this page |