comp.lang.idl-pvwave archive: archive » Re: 5th & 95th Percentiles

Home » Public Forums » archive » Re: 5th & 95th Percentiles

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: 5th & 95th Percentiles [message #58677]

Thu, 14 February 2008 01:04

Maarten[1]
Messages: 176
Registered: November 2005

Senior Member

On Feb 13, 5:04 pm, Spon <christoph.b...@gmail.com> wrote:

> pc5 = Data( (SORT(Data) ) (5 * N_ELEMENTS(Data) / 100) )
> pc95 = Data( (SORT(Data) ) (95 * N_ELEMENTS(Data) / 100) )

Please use

compile_opt defint32, strictarr

or

compile_opt idl2

at the beginning of each function (and in your IDL startup script) and
square braces for readability. For me at least it makes my head spin.
Consider the following instead:

pc5 = Data[ (SORT(Data))[5 * N_ELEMENTS(Data) / 100] ]

At least now it is clear which is which (array indexing vs. function
calling).

If you need both pc5 and pc95, be sure to store the sorted index:

sidx = sort(data)
ndata = n_elements(data)
pc5 = data[sixd[5*ndata / 100]]
pc66 = data[sixd[66*ndata / 100]]

Maarten

Report message to a moderator

[Message index]

		Re: 5th & 95th Percentiles By: Maarten[1] on Thu, 14 February 2008 01:04
		Re: 5th & 95th Percentiles By: Spon on Wed, 13 February 2008 09:04
		Re: 5th & 95th Percentiles By: chloesharrocks on Wed, 13 February 2008 08:19
		Re: 5th & 95th Percentiles By: Vince Hradil on Wed, 13 February 2008 08:10

Previous Topic:	workbench project rename: bug?
Next Topic:	Re: N_ELEMENTS gt 0

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 10 01:21:51 PDT 2025

Total time taken to generate the page: 0.80812 seconds