Repository navigation
plt.hist chooses improper range when using string-based bin options #8636
Description
plt.hist range bug
When using plt.hist with multiple input data sets, not supplying a range, and using a string-based binning option ('auto', for example), the range determined will cut off the later datasets and only fully include the first.
Code for reproduction
import numpy as np
import matplotlib.pyplot as plt
data = np.random.normal(0, 1, size=1000)
data2 = np.random.normal(3, 1, size=1000)
data3 = np.random.normal(6, 1, size=1000)
plt.hist([data, data2, data3], bins='auto', histtype='stepfilled', stacked=False)Actual outcome
Expected outcome
This is the plot using the bin-edges as determined by np.histogram using the following code:
_, be = np.histogram([data, data2, data3], bins='auto')
plt.hist([data, data2, data3], bins=be, histtype='stepfilled', stacked=False)
Matplotlib version
- Operating System: OSX 10.10.5
- Matplotlib Version: 2.0.1
- Python Version: 2.7.13
- Jupyter Version (if applicable): 4.0.6
- Other Libraries:
Python installed via Homebrew, all python packages installed via pip.
Activity
FIX: if bins input to hist is str, treat like no bins
tacaswell commentedon May 18, 2017
I have a PR to fix this, but am not super happy about it.
bebound commentedon Dec 13, 2017
Is this the same bug? Trying to draw a hist from Series, the second one is the expected output.
Thank you.
Source is https://www.kaggle.com/cast42/exploring-features , In [8]:
FIX: if bins input to hist is str, treat like no bins
FIX: if bins input to hist is str, treat like no bins
FIX: if bins input to hist is str, treat like no bins
FIX: if bins input to hist is str, treat like no bins