How to format bar chart values to two decimal places, only for the values present in the data?

Question

I am trying to plot a horizontal barchart using searborn. But I'd like the y-axis to display two decimal points of my data, but only for the values that are present in the data, for example 0.96, 0.93, ... .

Here is what I have:

df=pd.read_excel('file.xlsx', sheet_name='all')
print(df['digits'])

1.   0.96270
1    0.93870
2    0.93610
3    0.69610
4    0.61250
5    0.61280
6    0.52965
7    0.50520

sns.histplot(y=df['digits'])
plt.xlabel("frequency", fontsize=15)
plt.ylabel("results",fontsize=15)

Here is the output

I edited your question with the new information from your comments. — JohanC, Apr 12 '21 at 14:01

score 2 · Accepted Answer · answered Apr 13 '21 at 18:34

To create a histogram where the value rounded to 2 decimals defines the bin, you can create bin edges halfway between these values. E.g. bin edges at 0.195 and 0.205 would define the bin around 0.20. You can use `np.arange(-0.005, 1.01, 0.01)' to create an array with these bin edges.

In order to only set tick labels at the used positions, you can use ax.set_yticks(). You can round all the y-values and use the unique values for the y ticks.

If you don't want rounding, but truncation, you could use bins=np.arange(0, 1.01, 0.01) and ax.set_yticks(np.unique(np.round(y-0.005, 2))).

import numpy as np
import matplotlib.pyplot as plt
from matplotlib.ticker import MaxNLocator
import seaborn as sns

y = np.array([0.96270, 0.93870, 0.93610, 0.69610, 0.61250, 0.61280, 0.52965, 0.50520])

ax = sns.histplot(y=y, bins=np.arange(-0.005, 1.01, 0.01), color='crimson')
ax.set_yticks(np.unique(np.round(y, 2)))
ax.xaxis.set_major_locator(MaxNLocator(integer=True))
ax.tick_params(axis='y', labelsize=6)
ax.set_xlabel("frequency", fontsize=15)
ax.set_ylabel("results", fontsize=15)
plt.show()

Note that even with a small fontsize the tick labels can overlap.

Another approach, is to use a countplot on the rounded (or truncated) values. Then the bars get evenly spaced, without taking empty spots into account:

y = np.array([0.96270, 0.93870, 0.93610, 0.69610, 0.61250, 0.61280, 0.52965, 0.50520])
y_rounded = [f'{yi:.2f}' for yi in sorted(y)]
# y_truncated = [f'{yi - .005:.2f}' for yi in sorted(y)]
ax = sns.countplot(y=y_rounded, color='dodgerblue')

ax.xaxis.set_major_locator(MaxNLocator(integer=True))

Which approach did you prefer? – JohanC Apr 14 '21 at 09:43 — JohanC, Apr 14 '21 at 09:43

score 0 · Answer 2 · answered Apr 09 '21 at 16:00

0

This is handled by matplotlib:

ax = sns.histplot(y=np.random.randn(20))
ax.xaxis.set_major_formatter("{x:.2f}")
ax.set_xlabel("frequency", fontsize=15)
ax.set_ylabel("results",fontsize=15)

answered Apr 09 '21 at 16:00

mwaskom

46,693
16
125
127

Thank you @measkom, but i realy with to plot the realy value of my decimal points eg 0.96, 0.93 etc, is there any way of doing so? – TinaTz Apr 12 '21 at 12:03

score 0 · Answer 3 · edited Apr 09 '21 at 16:50

You might want to use this:

import matplotlib.ticker as tkr

y = [0.96270, 0.93870, 0.93610, 0.69610, 0.61250, 0.61280, 0.52965, 0.50520]
g = sns.histplot(y=y)
plt.xlabel("frequency", fontsize=15)
plt.ylabel("results",fontsize=15)

g.axes.yaxis.set_major_formatter(tkr.FuncFormatter(lambda y, p: f'{y:.2f}'))

set_major_formatter

or this:

import matplotlib.ticker as tkr

y = [0.96270, 0.93870, 0.93610, 0.69610, 0.61250, 0.61280, 0.52965, 0.50520]
g = sns.histplot(y=y, binwidth=0.01)
plt.xlabel("frequency", fontsize=15)
plt.ylabel("results",fontsize=15)

g.axes.yaxis.set_major_formatter(tkr.FuncFormatter(lambda y, p: f'{y:.2f}'))

binwidth=0.01:

binwidth=0.01

Thank you very much @JohanC for this answer, But I want to plot the real value, for example 0.96, 0.93, what should I do? — TinaTz, Apr 12 '21 at 12:02

How to format bar chart values to two decimal places, only for the values present in the data?

3 Answers3