I am trying to plot a histogram of my data, and I seem to be a little confused here. I am using matplotlib in Python. Here is the code from their website:

mu = 100 #mean
sigma = 15 #std deviation
x = mu + sigma * np.random.randn(10000)

# the histogram of the data
n, bins, patches = plt.hist(x, num_bins, normed=1, facecolor='green', alpha=0.5)

I am confused as to what the x-axis should be for my use. I have calculated the standard deviation and the mean, but I am uncertain if I should replace the np.random.randn(10000) with the actual data.


It sounds like what you want to do is completely replace x in the plotting function with your data, so what you should get looks like this:

n, bins, patches = plt.hist(my_data, num_bins, normed=1, facecolor='green', alpha=0.5)

Also, if you aren't planning on using n, bins, or patches you can discard them by just running:

plt.hist(my_data, num_bins, normed=1, facecolor='green', alpha=0.5)

If this isn't correct please let me know and I'll be happy to help.

  • $\begingroup$ Did everything work out for you? $\endgroup$ – Henry Hammond Jul 31 '14 at 19:13
  • $\begingroup$ yes, I am able to interpret my data now. Just one last question, when I plot this, what will go on the y-axis? It looks like it may be what portion of my data a certain bin will fall in. $\endgroup$ – alex Jul 31 '14 at 19:15
  • $\begingroup$ Because you have normed=1 matplotlib will normalize the y-axis to represent what percentage of your data fell in that bin. If you want the counts for each bin you can simply remove the whole normed argument. $\endgroup$ – Henry Hammond Jul 31 '14 at 19:16
  • $\begingroup$ I see. this makes sense. My data is just looking odd. Median ~29 and mean~119 $\endgroup$ – alex Jul 31 '14 at 19:22
  • $\begingroup$ I looked into it and my above comment is incorrect. From the Numpy docs > If True, the result is the value of the probability density function at the bin, normalized such that the integral over the range is 1 $\endgroup$ – Henry Hammond Jul 31 '14 at 19:24

The code you proposed for generating random numbers will not generate a distribution centered on mean and with a standard deviation sigma, as your variable names suggest. Note that if you calculate using


you'll get something around 0.28 and the reason is simple: np.random.randn(k) draws k numbers between 0 and 1. To generate a vector with 10 000 numbers following a gaussian distribution of parameters mu and sigma use

from random import gauss
x=[gauss(mu, sigma) for i in range(10000)]

for which in the last line I used the "pythonic" condensed version of a for loop, the list comprehension. Then you can feed your x vector to the histogram plotting routine, that will calculate the histogram of a vector for plotting.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.