Histogram Worksheet - so sloooooow

[JBryant61-disab]

Hi, I am trying to use a histogram as a means to measure image uniformity. To do this I first down sample the raw image to a binned version (this is the tabled data in the worksheet). I then partition this downsampled image into a histogram and use this for a calculation. As my images can have a widely differing intensity profiles, I want to keep the number of bins fixed rather than bin with an absolute value of intensity. I thus normalise the bin.

The whole worksheet is extremely slow

Is there anything I can do to speed it up.

Thanks

Jason

BTW, I have finally converted to mathcad 15!

[StuartBruff]

Could you please post the worksheet in an earlier format as well? M11 preferably, M13 if not.

Stuart

[JBryant61-disab]

Here You Go

[StuartBruff]

Jason,

I've just opened your worksheet and the first thing I notice is that you actually calculate 800+ bins for the histograms P & P1 at the start of the collapsed area rather than the 20 the text states ... I'm not sure that you are caculating the correct thing, looking at .

Stuart

[StuartBruff · Accepted solution]

Jason,

does the attached do something like what you want?

The histogram function requires that you either provide the number of bins for equal spacing (which is what I think you want) or a vector of bin end points (nbins+1 points). I think all you need to do is defined nbins:=20 and the define P:=histogram(nbins,Z), etc.

Stuart

Message was edited by: Stuart_Bruff - worksheet corrected to create P1 from Z1 rather than Z

[JBryant61-disab]

Stuart, Thanks for pointing out my error. Its much faster now I am calculating the correct numbe rof bins.

Jason