[FIXED] How to limit x axis to non-empty region of histogram with predetermined bins

Issue

I have this data (dfe) that I want to plot in a histogram

Nr        O18ad
0        -4.268475
1        -4.265793
2        -4.263120
3        -4.260457
4        -4.257803
            ...
359995   -7.813345
359996   -7.821394
359997   -7.773479
359998   -7.807605
359999   -7.797769

Here I have this code snippet: It is part of an 8 part figure hence the subfigure function

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from scipy.stats import norm

def plot_Diff(data):
    fig = plt.figure(figsize=(10, 10))
    ax =fig.add_subplot(423)  
    x= dfe['O18ad']
    bins=[-20, -19, -18, -17, -16, -15, -14, -13, -12, -11, -10, -9, -8, -7, -6, -5, -4, -3, -2, -1, 0, 1]
    old_yticks = ax.get_yticks()
    bin_counts, _, bars = plt.hist(x, bins,  alpha=0.65, label='old', edgecolor='black', color='lightgrey')
    new_max_tick = bin_counts.max()
    old_yticks = ax.get_yticks()
    new_ticks =  old_yticks[old_yticks < new_max_tick][:-1]
    new_ticks = np.append(new_ticks, new_max_tick)
    ax.set_yticks(new_ticks)
    manual_ticks=[0,20_000]
    if manual_ticks is None:
        old_yticks = ax.get_yticks()
        new_ticks = old_yticks[old_yticks < new_max_tick][:-1] 
        new_ticks = np.append(new_ticks, new_max_tick)
    else:
        new_ticks = np.append(manual_ticks, new_max_tick)
    ax.set_yticks(new_ticks)

plt.show()

My output looks like this:

Now my question is how can I change the X axis ticks so that the maximum values (in this example 0) are automatically set in the right corner and the displayed minimum (in this example -20) values are in the left corner?

It is important that the max and min values are set automatically because I have about 100 histograms where I cannot always set them manually.

Solution

If I understand correctly, you want to find the left edge of the leftmost non-empty bin, and similarly at the right.

When bins is a numpy array of bin edges, the left edges are bins[:-1] (first till next to last edge), while the right edges are bins[1:](second till last edge). As bin_counts is also a numpy array, you can index with [bin_counts > 0] to filter out the non-empty bins. The obtained min and max will serve as the axis limits.

import matplotlib.pyplot as plt
import pandas as pd
import numpy as np

dfe = pd.DataFrame({'O18ad': np.random.normal(0.01, 0.1, 1000).cumsum()})
dfe['O18ad'] -= dfe['O18ad'].max() + np.random.uniform(0.1, 5)

fig = plt.figure(figsize=(10, 10))
ax = fig.add_subplot(111)
x = dfe['O18ad']
bins = np.arange(-20, 2)
old_yticks = ax.get_yticks()
bin_counts, _, bars = plt.hist(x, bins, alpha=0.65, label='old', edgecolor='black', color='lightgrey')
new_max_tick = bin_counts.max()
print(bins[:-1][bin_counts > 0].min())
print(bins[1:][bin_counts > 0].max())

ax.set_xlim(bins[:-1][bin_counts > 0].min(), bins[1:][bin_counts > 0].max())

plt.show()

Answered By - JohanC