PHS about hilo_paper

[petrelharp]

It just seemed a bit odd to me to call it a p-value: usually "p-value" is computed under a statistical model, or permutations, or something; not just a quantile in the original data. Maybe I'm thinking about something wrong, and this is in fact standard.

p.s. If you click on source code to get a link to the line number, like you linked to above, you can then hit 'y' to get a canonical link, including the commit, like this one that won't change when the source code changes.

[rossibarra]

Whoa. Cool! Sorry I didn't see your point -- yes not really a p-value.
Have a suggested wording?

---

Jeffrey Ross-Ibarra

Dept. of Plant Sciences
262 Robbins Hall, Mail Stop 4
University of California
One Shields Ave
Davis, CA 95616

Tel: [phone]
@jrossibarra
www.rilab.org

On Fri, Dec 26, 2014 at 9:11 PM, Peter Ralph [email protected]
wrote:

It just seemed a bit odd to me to call it a p-value: usually "p-value" is
computed under a statistical model, or permutations, or something; not just
a quantile in the original data. Maybe I'm thinking about something wrong,
and this is in fact standard.

p.s. If you click on source code to get a link to the line number, like you
linked to above
https://github.com/rossibarra/hilo_paper/blob/master/methods.tex#L177,
you can then hit 'y' to get a canonical link, including the commit, like this
one

hilo_paper/methods.tex

Line 177 in 4f15af4

%\plr{This is (one minus) the quantile of the PHS value.}

that won't change when the source code changes.

—
Reply to this email directly or view it on GitHub
#11 (comment)
.

[petrelharp]

How about replace "p-value" with "quantile" and leave the sentence mostly
as is, but say "To identify outlying PHS values, we used the empirical
quantile, calculated as the proportion..."

On Fri, Dec 26, 2014 at 10:47 PM, Jeffrey Ross-Ibarra <
[email protected]> wrote:

Whoa. Cool! Sorry I didn't see your point -- yes not really a p-value.
Have a suggested wording?

---

Jeffrey Ross-Ibarra

Dept. of Plant Sciences
262 Robbins Hall, Mail Stop 4
University of California
One Shields Ave
Davis, CA 95616

Tel: [phone]
@jrossibarra
www.rilab.org

On Fri, Dec 26, 2014 at 9:11 PM, Peter Ralph [email protected]
wrote:

It just seemed a bit odd to me to call it a p-value: usually "p-value"
is
computed under a statistical model, or permutations, or something; not
just
a quantile in the original data. Maybe I'm thinking about something
wrong,
and this is in fact standard.

p.s. If you click on source code to get a link to the line number, like
you
linked to above
https://github.com/rossibarra/hilo_paper/blob/master/methods.tex#L177,

you can then hit 'y' to get a canonical link, including the commit, like
this
one
<
https://github.com/rossibarra/hilo_paper/blob/4f15af47b34a78cf5addb05ddedf760b6098b5e2/methods.tex#L177>

that won't change when the source code changes.

—
Reply to this email directly or view it on GitHub
<
https://github.com/rossibarra/hilo_paper/issues/11#issuecomment-68169135>
.

—
Reply to this email directly or view it on GitHub
#11 (comment)
.