具有非规范拟合线的直方图，matplotlib

Question

I'm trying to get a set of histograms plotted, with raw count data (non-normalized to density/pdf) and a fit line. However, I can't seem to figure out how to get a fit line plotted that ISN'T normalized by a pdf function. Is there a way to plot a non-normalized line, or a function to reverse the density calculation? Right now, I've got the below code, which works for the normalized histogram and fit line.

fig, ax = plt.subplots()
x=[13.140,17.520,15.768,10.512,10.512,9.636,10.512, 9.636,11.388,7.884,7.008,7.008,9.636,11.388,7.884,7.88,16.64‌​4,42.924,17.520]

n, bins, patches = plt.hist(x, bins=10, normed=False, color='cornflowerblue', alpha=0.75)
(mu, sigma) = norm.fit(x)
y = mlab.normpdf(bins, mu, sigma)
l = plt.plot(bins, y, '-o', linewidth=2)
ax.set_xlabel('Millirems')

This is the graph i have so far, with raw count data and a normalized fit line

Answer 1

You could just do this by multiplying the pdf by the total area of the histogram I think?

import numpy as np

l = plt.plot(bins, y * np.sum(np.diff(bins) * n))

Answer 2

Maybe you want to scale the pdf by the same factor the histogram is scaled with respect to a normed one. This factor would be the area of the histogram sum(n * np.diff(bins)) .

fig, ax = plt.subplots()
x = [13.140,17.520,15.768,10.512,10.512,9.636,10.512, 9.636,11.388,7.884,7.008,7.008,9.636,11.388,7.884,7.88,16.644,42.924,17.520]

n, bins, patches = plt.hist(x, bins=10, normed=False, color='cornflowerblue', alpha=0.75)
(mu, sigma) = norm.fit(x)
y = mlab.normpdf(bins, mu, sigma) * sum(n * np.diff(bins))
plt.plot(bins, y, '-o', linewidth=2)
ax.set_xlabel('Millirems')