Plot the peaks of a histogram

Question

I'm trying to figure out how to plot the peaks of a simple histogram using scipy.signal.find_peaks but the peaks found seem way off.

ages = np.array([10, 5, 22, 13, 50, 45, 67, 30, 21, 34, 60, 67, 89, 45, 45, 65])
hist, bin_edges = np.histogram(ages, 10)
bin_edges = bin_edges[1:]

plt.plot(bin_edges, hist)
peaks, _ = find_peaks(hist)
plt.plot(ages[peaks], peaks, "x")

sim0ne · Accepted Answer · 2019-07-30 18:59:24Z

4

You should try:

plt.plot(bin_edges[peaks], hist[peaks], "x")

find_peaks gives you the indices of local maxima in the hist signal.

The x-values of your histogram are bin_edges and the y-values are given by hist. So you have to look for the indices given by peaks in each of these series.

edited Jul 30, 2019 at 18:59

answered Jul 30, 2019 at 18:26

sim0ne

3221 silver badge7 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Brien Crean Over a year ago

Wow that looks great! Thank you! That looks like the right answer to me. Can you add an explanation as to what this line is doing? How is it relating the peaks to the x and y since peaks is a 1d array?

sim0ne Over a year ago

I updated the answer with a bit more explanation :)

Dai Over a year ago

It is worth noting that histograms are not always the best way to visualise a distribution, and can lead to misleading interpretations of the data: a simple change of bin width can drastically change their form. Depending on your application, a better(?) way may be to calculate the "kernel density estimation" (see jakevdp.github.io/blog/2013/12/01/kernel-density-estimation) though, of course, in this example it might be overkill!

Collectives™ on Stack Overflow

Plot the peaks of a histogram

1 Answer 1

3 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

3 Comments

Linked

Related