How to get the data as output along with the histogram plot?

Hi,

I have a set of data which looks like

0.499783

0.499669

0.500231

0.499909

0.499542

0.50018

0.500785

0.49823

.... and so on (say I haev 50000 rows)

Now I am plotting

histogram(S,'Normalization','probability') , where S is the column containing the data.

I get the normalized probability histogram as a plot. But I want the data that is finally plotted, i.e., no. of hits in each bin or something like that. Such that in future I can get the same histogram plot without using the original raw data. Is that possible?

Answer 1

[Tommy]

You could request output from histogram():

h = histogram(S,'Normalization','probability');

and access properties such as h.BinEdges, h.BinCounts, h.Values, etc, described here.

Or you could use histcounts(), described here, after you plot your histogram:

histogram(S,'Normalization','probability');
[binCounts,binEdges] = histcounts(S,'Normalization','probability');

[Angana Ray]

Hi Tommy,

Thank you very much for the answer. The binCounts and binEdges is what I needed. Now, my query is say I have only the binCounts and BinEdges and do not have the original raw data, how can I plot the histogram. My question is how can I replicate the same histogram plot I had when I used histogram(S,'Normalization','probability')

Thank you.

[Tommy]

Happy to help!

Easiest I think would be to use the following syntax of histogram():

histogram('BinEdges', binEdges, 'BinCounts', binCounts)

or you can try to plot it yourself using bar().

[Angana Ray]

Can you please elaborate bar() ?

Thank you.

[Tommy]

A histogram is basically a bunch of adjacent bars, with one bar per bin. So you could determine the centers of each bin and then call bar(), using a width of 1 to avoid any gap between the bins:

binCenters = binEdges(1:end-1) + diff(binEdges)/2;
bar(binCenters, binCounts, 1)

By adjusting the color and x limits (and possibly other properties) you could probably replicate the histogram plot exactly.

Using

histogram('BinEdges', binEdges, 'BinCounts', binCounts)

would definitely create the exact same plot (or at least I assume it would).

[Angana Ray]

Thank you very very much