R help - Mar 2013 - Sen's slope - fume package different output than zyp or wq

Hello,

I am trying to decide which package to use to calculate the non-parametric
Sen's Slope for identifying trends in rainfall data (determine the slope
between all pairs of points and take the median of those slopes).  I have
found three packages that output Sen's: "zyp", "wq" and
"fume".  The
outputs of "zyp.sen()" and "mannKen()" from the zyp and wq
packages match,
but the output from fume's "mkTrend" is different.  I tried
looking at the
code from these three and I could not really understand what zyp and wq are
doing differently.

It seems like "fume" is the only package that calculates a corrected
Mann
Kendall p-value based on temporal autocorrelation, so I would like to use
this package to calculate my p-values, but I am cautious because if the
calculations for Sen's are incorrect, maybe the corrected p-value is not
correct either?

Please let me know if I should provide some examples - but the
discrepancies seem to show up in any dataset I use.  Thanks!

Abby

	[[alternative HTML version deleted]]

On 22.03.2013 01:09, Abby Frazier wrote:
> Hello,
>
> I am trying to decide which package to use to calculate the non-parametric
> Sen's Slope for identifying trends in rainfall data (determine the
slope
> between all pairs of points and take the median of those slopes).  I have
> found three packages that output Sen's: "zyp", "wq"
and "fume".  The
> outputs of "zyp.sen()" and "mannKen()" from the zyp and
wq packages match,
> but the output from fume's "mkTrend" is different.  I tried
looking at the
> code from these three and I could not really understand what zyp and wq are
> doing differently.
>
> It seems like "fume" is the only package that calculates a
corrected Mann
> Kendall p-value based on temporal autocorrelation, so I would like to use
> this package to calculate my p-values, but I am cautious because if the
> calculations for Sen's are incorrect, maybe the corrected p-value is
not
> correct either?
>
> Please let me know if I should provide some examples - but the
> discrepancies seem to show up in any dataset I use.  Thanks!

I'd start to discuss inaccuracies or errors you found with the package 
maintainer in in case your are still in doubt look into the source code.

Best,
Uwe Ligges
>
> Abby
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>