| Re: Non-linear data selection [message #70996] |
Sun, 16 May 2010 06:16 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Gianguido Cianci writes:
>
> On May 15, 7:27 pm, David Fanning <n...@dfanning.com> wrote:
>> aboveIndices = Where((y GE yfit) EQ 1, count)
>
> David, why didn't you use one of the following?
>
> aboveIndices2 = Where(y GE yfit, count)
> aboveIndices3 = Where(y[*] GE yfit[*], count)
>
> I have this vague memory/nightmare that I "need" to do the [*] part in
> order not to get crazy results sometimes (though I can't come up with
> a crazy example right now) so I always use aboveIndices3...
>
> So *are* there differences among these three lines? Cos they all seem
> to work for the example of the original poster. :-|
The second line (aboveIndices2) is what I originally
wrote in my code, but I changed it to give more clues
to the user about what I was doing. (I was too busy working
on my rock garden yesterday to explain myself.)
The third line is totally unnecessary and, at least with
large arrays, will cost you a big memory penalty:
http://www.dfanning.com/misc_tips/submemory.html
But, yes, they should all be equivalent in this case.
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
| Re: Non-linear data selection [message #70997 is a reply to message #70996] |
Sun, 16 May 2010 00:18  |
cgguido
Messages: 195
Registered: August 2005
|
Senior Member |
|
|
On May 15, 7:27 pm, David Fanning <n...@dfanning.com> wrote:
> aboveIndices = Where((y GE yfit) EQ 1, count)
David, why didn't you use one of the following?
aboveIndices2 = Where(y GE yfit, count)
aboveIndices3 = Where(y[*] GE yfit[*], count)
I have this vague memory/nightmare that I "need" to do the [*] part in
order not to get crazy results sometimes (though I can't come up with
a crazy example right now) so I always use aboveIndices3...
So *are* there differences among these three lines? Cos they all seem
to work for the example of the original poster. :-|
Thanks,
Gianguido
|
|
|
|
|
|
| Re: Non-linear data selection [message #71000 is a reply to message #70998] |
Sat, 15 May 2010 17:27 |
David Fanning
Messages: 11724
Registered: August 2001
|
Senior Member |
|
|
Joe Daal writes:
> There is an interesting problem that I often come across.....
> I have a scattered dataset of x and y, say 1000 elements each. The
> data could be of any shape, for now, let's assume it is scattered
> allover the X-Y window.
>
> There is straight line (y=mx+b, could be some other curve as well)
> that that crosses the plot, I need to select all data points this
> line.
> If it is just a one value for y, the WHERE statement is ok, but how to
> do this?
x = Randomu(-3L, 100) * 10
y = Randomu(-5L, 100) * 10
m = 0.5
b = 3.25
yfit = m*x+b
aboveIndices = Where((y GE yfit) EQ 1, count)
Print, StrTrim(count,2) + ' points lie above the line.'
Plot, x, y, /NoData, Background=FSC_Color('white')
OPlot, x, y, PSYM=2, SYMSIZE=0.5, Color=FSC_Color('sky blue')
OPLOT, x[aboveIndices], y[aboveIndices], PSYM=4, Color=FSC_Color('red')
OPlot, x, yfit, LINESTYLE=2
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting, Inc.
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Sepore ma de ni thui. ("Perhaps thou speakest truth.")
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
