Login
Password
Sources on this Page

> Headlines by Category

 Home / Science / Mathematics

You are using the plain HTML view, switch to advanced view for a more complete experience.

findInRange

When constructing peri-event time histograms and raster-plots it's often necessary to repeatedly do the following operation:

find(X > T1 & X <= T2)

where X is a vector of the times of some measurements (eg a spike train) and T1 and T2 are the left and right edges of an event-aligned window (eg stimulus onset).

While find() is generally fast, it can get quite slow for large numbers of events/trials because it is necessary to loop over the vector of alignment-events. findInRange() uses mex to achieve much greater speeds:

X = rand(1e6,1)*1e6;
T1 = rand(1e3,1)*1e3;
T2 = T1 + 5;

tic;
V1 = findInRange(X, T1, T2);
time1 = toc;

tic;
for i=1:length(T1)
    V2{i} = X(X > T1(i) & X <= T2(i));
end
time2 = toc;

time2/time1

ans =

  112.7594

Re: help with graphing sin x
ezplot(@(x) sinc(x/pi),[-30,30])
ylim([-.4,1.5])
difference approximation with diff
Hi,

I've a vector meanfield size 360x480.I want to do difference approximation using command diff. When operate using this command, the output size become 359x480.
So, I'm unable to proceed next operation since the size of the vector has been changed. Is it right to use diff command for difference approximation? Is there any solution to overcome this problem?

h = 0.023; % step size
dMeanField = diff(meanfield);

I want to find dmeanfield/h.

Thanks.
Re: profile connection in appearance matching
On Tue, 16 Sep 2014 15:37:30 -0400, Dale <invalid@invalid.invalid>
wrote:

>I'm talking in terms of the icc, http://www.color.org/
>but it could apply to any device independent color system
>
>if you talk about "color" matching, it seems to me that absolute
>colormetry should be the correct approach, this is simply measured and
>a simple cmm can connect device colors to and from the absolute and
>static profile connection space
>
>if you are talking about "appearance" matching, an absolute color
>profile connection space and a simple cmm do not apply I think, each
>device and ideally its device profile have its own appearance
>considerations and the cmm has to connect the "appearances" through
>the profile connection space with those "appearance" considerations
>
>it would seem to me that the correct approach is for the cmm to
>instantiate a profile connection space based on the input and output
>"appearance" considerations at the time of connection,
>
>the instantiation could be an event-loop or an OO approach or
>whatever, but according to CIECAM02, LMS color space is involved
>http://en.wikipedia.org/wiki/CIECAM02

I notice icc has rimm and romm (reference input/output) medium metrics
that have components of appearance similiar to CIECAM02, but they
don't have a reference for display

I also gather the reference spaces are a high gamut film

shouldn't they just get rid of the references and go with a full
appearance models that instaniates a profile connection space based on
i/i paramaters, should this requires a more complex measurement
paradigm, and maybe "good enough" is "good enough"

just a thought

--
Dale
Re: output device profile rendering intents
On Tue, 16 Sep 2014 21:54:51 -0400, Dale <invalid@invalid.invalid>
wrote:

>in the icc http://www.color.org a output device profile has founr
>possible rendering intents, I think
>
>1) absolute colorimteric, measurement only basically
>
>2) relative colorimetric, an attempt at a general appearance model
>based on white point translation
>
>3) hold saturation, where smaller saturation changes are preferred
>over whole colorimetric changes in things like business presentation
>displays
>
>4) perceptual, an appearance model based on an ideal print, I think
>
>why not change 4 into a selection of possible output appearance
>models? instead of just one, the print

I notice icc has rimm and romm (reference input/output) medium metrics
that have components of appearance similiar to CIECAM02, but they
don't have a reference for display

I also gather the reference spaces are a high gamut film

shouldn't they just get rid of the references and go with a full
appearance models that instaniates a profile connection space based on
i/i paramaters, should this requires a more complex measurement
paradigm, and maybe "good enough" is "good enough"

just a thought
--
Dale
output device profile rendering intents
in the icc http://www.color.org a output device profile has founr
possible rendering intents, I think

1) absolute colorimteric, measurement only basically

2) relative colorimetric, an attempt at a general appearance model
based on white point translation

3) hold saturation, where smaller saturation changes are preferred
over whole colorimetric changes in things like business presentation
displays

4) perceptual, an appearance model based on an ideal print, I think

why not change 4 into a selection of possible output appearance
models? instead of just one, the print

--
Dale
Re: help with graphing sin x
On Wednesday, September 17, 2014 1:02:10 PM UTC+12, Stephanie wrote:
> I have to graph y=sin(x)/x from the interval -30 to 30 and it just will not let me use this equation because it is not defined. Does anyone know what to do?!

Google L'Hospital
help with graphing sin x
I have to graph y=sin(x)/x from the interval -30 to 30 and it just will not let me use this equation because it is not defined. Does anyone know what to do?!
Post Selected Items to:

Showing 10 items of about 1200

home  •   advertising  •   terms of service  •   privacy  •   about us  •   contact us  •   press release design by Popshop •   Official PR partner B2BLogger.com •   © 1999-2014 NewsKnowledge