Something strange in cor.test in R-1.8.0 (PR#4718)

Without knowing the seed used it is impossible for us to reproduce this,
but I am not seeing anything strange.

Could you please reply with a reproducible example (and list the data used 
too, in case the random number generation is the issue).

Full_Name: Ian Wilson
Version: R-1.8.0
OS: Windows (but own compilation)
Submission from: (NULL) (139.133.7.38)



the p-value is incorrect for cor.test using method "spearman" in R-1.8.0.  This
was not the case in R-1.7.1.

Version R-1.8.0 on Windows 

cor.test(rnorm(50),rnorm(50),method="spearman")

	Spearman's rank correlation rho

data:  rnorm(50) and rnorm(50) 
S = 20486, p-value = < 2.2e-16
alternative hypothesis: true rho is not equal to 0 
sample estimates:
       rho 
0.01627851 

cor.test(rnorm(50),rnorm(50),method="spearman")

	Spearman's rank correlation rho

data:  rnorm(50) and rnorm(50) 
S = 22006, p-value = 1
alternative hypothesis: true rho is not equal to 0 
sample estimates:
        rho 
-0.05671068

______________________________________________
R-devel@stat.math.ethz.ch mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-devel

Brian D. Ripley,                  ripley@stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

Without knowing the seed used it is impossible for us to reproduce this,
but I am not seeing anything strange.

I'm getting strange results in 1.8.0 for Windows too.

set.seed(1)
x <- rnorm(50)
y <- rnorm(50)
cor.test(x,y,method="spearman")

Spearman's rank correlation rho

data:  x and y 
S = 23640, p-value = 1
alternative hypothesis: true rho is not equal to 0 
sample estimates:
       rho 
-0.1351741

x <- rnorm(50)
y <- rnorm(50)
cor.test(x,y,method="spearman")

Spearman's rank correlation rho

data:  x and y 
S = 17248, p-value = < 2.2e-16
alternative hypothesis: true rho is not equal to 0 
sample estimates:
      rho 
0.1717647

x

[1] -0.62036668  0.04211587 -0.91092165  0.15802877 -0.65458464
1.76728727
 [7]  0.71670748  0.91017423  0.38418536  1.68217608 -0.63573645
-0.46164473
[13]  1.43228224 -0.65069635 -0.20738074 -0.39280793 -0.31999287
-0.27911330
[19]  0.49418833 -0.17733048 -0.50595746  1.34303883 -0.21457941
-0.17955653
[25] -0.10019074  0.71266631 -0.07356440 -0.03763417 -0.68166048
-0.32427027
[31]  0.06016044 -0.58889449  0.53149619 -1.51839408  0.30655786
-1.53644982
[37] -0.30097613 -0.52827990 -0.65209478 -0.05689678 -1.91435943
1.17658331
[43] -1.66497244 -0.46353040 -1.11592011 -0.75081900  2.08716655
0.01739562
[49] -1.28630053 -1.64060553

y

[1]  0.45018710 -0.01855983 -0.31806837 -0.92936215 -1.48746031
-1.07519230
 [7]  1.00002880 -0.62126669 -1.38442685  1.86929062  0.42510038
-0.23864710
[13]  1.05848305  0.88642265 -0.61924305  2.20610246 -0.25502703
-1.42449465
[19] -0.14439960  0.20753834  2.30797840  0.10580237  0.45699881
-0.07715294
[25] -0.33400084 -0.03472603  0.78763961  2.07524501  1.02739244
1.20790840
[31] -1.23132342  0.98389557  0.21992480 -1.46725003  0.52102274
-0.15875460
[37]  1.46458731 -0.76608200 -0.43021175 -0.92610950 -0.17710396
0.40201178
[43] -0.73174817  0.83037317 -1.20808279 -1.04798441  1.44115771
-1.01584747
[49]  0.41197471 -0.38107605


In 1.7.1, the same code gives the same x and y and rho values, but
more reasonable p-values:

cor.test(x,y,method="spearman")

Spearman's rank correlation rho

data:  x and y 
S = 23640, p-value = 0.3482
alternative hypothesis: true rho is not equal to 0 
sample estimates:
       rho 
-0.1351741

x <- rnorm(50)
y <- rnorm(50)
cor.test(x,y,method="spearman")

Spearman's rank correlation rho

data:  x and y 
S = 17248, p-value = 0.2322
alternative hypothesis: true rho is not equal to 0 
sample estimates:
      rho 
0.1717647

I've now tried the code 

set.seed(1)
x <- rnorm(50)
y <- rnorm(50)
cor.test(x,y,method="spearman")
x <- rnorm(50)
y <- rnorm(50)
cor.test(x,y,method="spearman")

in r-patched (which gives apparently correct results), and r-devel,
which gives the first result correct, the second incorrect; when
repeated, both are bad.

Duncan Murdoch