Re: Frequency Dist

[Steve Cocks <scocks@xxxxxxxxxxxxxxxxxx>]

Thanks Billy,

I was busy working with the original program, trying to find where the
error was.  However, your script worked just fine.  Thanks again.

Scocks

Billy Kessler wrote:

> Hi Scocks -
>
> You are right, frequency_histogram.jnl appears to be
> broken, or at least has not been updated since 1994.
>
> However, thanks to Gabe Vecchi, I recently learned
> how to make histograms. The following assumes that
> your time series has a free z axis. If not, then you
> will need to write out your time series to a dummy
> file, and read it back in as a plain time axis.
> ! ------------------------------------------------
> ! make a new z-axis that encompasses the range of the data
> define axis/z=-10:10:1 zaz
> define grid/z=zaz gaz
>
> ! choose histogram intervals (here .5)
> let inter = if (abs(timeseries-z[gz=gaz]) le .5 then 1
> let histo = inter[l=@sum]
>
> ! histo is now a count of the number of occurrences
> ! in bins of +-.5 units of zaz
> plot histo
>
> ! an obvious extension to plot percentages is
> plot 100*histo/`histo[k=1:21@sum]`
> ! ------------------------------------------------
> Regards .... Billy K