Plot x-axis with string array as in the same order in original array and not sort it alphabetically in matplotlib

Question

Either Numpy or Matplotlib is changing the order of my np.array and it's conflicting with my plot. It's causing the months to be out of order while the corresponding data to still be in the same order which is causing the plot to look weird:

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

f = np.array([53, 56, 63, 72, 79, 86, 89, 88, 83, 74, 65, 56])
month = np.array(["Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"])
plt.plot(month, f)

plt.xlabel('Month')
plt.ylabel('Temperature')
plt.title('Average Monthly Temperature in Elizabeth City, NC')
plt.show()

This is what i get as output in JupyterNotebook:

Oh okay. It's weird that its still using the corresponding data in the right order. — Edsel Norwood
– Edsel Norwood, Commented Jan 24, 2018 at 0:50
Sorry, what I meant to say is: matplotlib is turning the strings into x-coordinates by ordering all the strings, and returning the position of each one. The order of the line is unchanged, and still passes through the months in order — Eric
– Eric, Commented Jan 24, 2018 at 1:40

kmario23 · Accepted Answer · 2018-01-24 01:37:47Z

6

Since month is a string array, plt.plot() command is sorting it alphabetically. So, we have to use the xticks and then plot it like below to get the strings in the same order as it were in the original array month.

In [16]: f = np.array([53, 56, 63, 72, 79, 86, 89, 88, 83, 74, 65, 56])
    ...: month = np.array(["Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"])
    ...: plt.xticks(range(len(f)), month)
    ...: plt.plot(f)

Plot:

Note: For more customized plots refer: pylab date demo

edited Jan 24, 2018 at 1:37

answered Jan 24, 2018 at 1:17

kmario23

62k17 gold badges173 silver badges159 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Edsel Norwood Over a year ago

Thanks a bunch. I truly appreciate it :)

Eric Over a year ago

Note that this won't make the month of january wider than february to reflect its extra days, unlike @AGN's solution below

kmario23 Over a year ago

@Eric I've added a note at the bottom!

AGN Gazer · Accepted Answer · 2018-01-24 01:53:41Z

1

You will need to use MonthLocator and set_major_locator as shown here: formatting timeseries x-axis in pandas/matplotlib

Here is my attempt:

import matplotlib.pyplot as plt
import numpy as np
import datetime
f = np.array([53, 56, 63, 72, 79, 86, 89, 88, 83, 74, 65, 56])

# New stuff:
from matplotlib.dates import MonthLocator, DateFormatter
dates = []
for month in range(1, 13):
    dates.append(datetime.datetime(year=2018, month=month, day=1))
plt.plot(dates, f)
ax = plt.gca()
ax.set_xlim([dates[0], dates[-1]])
ax.xaxis.set_major_locator(MonthLocator())
ax.xaxis.set_major_formatter(DateFormatter('%b'))

plt.xlabel('Month')
plt.ylabel('Temperature')
plt.title('Average Monthly Temperature in Elizabeth City, NC')
plt.show()

edited Jan 24, 2018 at 1:53

answered Jan 24, 2018 at 0:50

AGN Gazer

8,4272 gold badges31 silver badges49 bronze badges

3 Comments

Edsel Norwood Over a year ago

Im having trouble trying to implement those with my existing code

AGN Gazer Over a year ago

@EdselNorwood I have added an example to my answer. Check it out.

Eric Over a year ago

Once you have an ax object, you may as well call all the plt methods on it too - ax.xlabel, ax.ylabel, and ax.title

Collectives™ on Stack Overflow

Plot x-axis with string array as in the same order in original array and not sort it alphabetically in matplotlib

2 Answers 2

3 Comments

3 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

3 Comments

Linked

Related