CODEWITHSUNDEEP
pythonmatplotlib
Wikunia
Wikunia

Reputation: 1592

Missing values at the beginning - matplotlib

Trying to add missing values (None) or in numpy np.nan to my matplot graph. Normally it is possible to have missing values in this way:

import matplotlib.pyplot as plt

fig = plt.figure(1)

x = range(10)
y = [1, 2, None, 0, 1, 2, 3, 3, 5, 10]

ax = fig.add_subplot(111)
ax.plot(x, y, 'ro')

plt.show()

But looks like missing values aren't easily supported at the beginning of the graph. 

import matplotlib.pyplot as plt

fig = plt.figure(1)

x = range(10)
y = ([None]*5)+[x for x in range(5)]

ax = fig.add_subplot(111)
ax.plot(x, y, 'ro')

plt.show()

This displays:

But I would like to start with x=0 instead of x=5. Hope there is a simple method for this problem.

Upvotes: 2

Views: 1895

Answers (2)

Suever
Suever

Reputation: 65430

The reason why the empty values don't seem to have an effect on the second plot is because of the automatic determination of the x and y limits of the axes object. In the first case, you have "real" data that spans the whole x / y range you care about and the axes is automatically scaled to fit this data. In the second example, your real data only covers the upper range of x values so matplotlib will truncate the x axis because there is nothing to show.

The way around this is to explicitly specify the xlims of the axes. You can use your value for x to determine what these should be. 

plt.xlim(x[0], x[-1])

Upvotes: 4

Nikolas Rieble
Nikolas Rieble

Reputation: 2601

Matplotlib automatically choses the parameters for x- and y- axes. In this case it is only showing the dots that can be visualized. If either x- or y is "None", nothing can be visualized. If you yet want to see the "empty" part of the graph, just adjust the axis range yourself.

How to do so, has been answered multiple times in similiar questions in this forum such as:

Matplotlib/pyplot: How to enforce axis range?

Python, Matplotlib, subplot: How to set the axis range?

setting y-axis limit in matplotlib

and plenty more. Therefore I will refrain from repeating what has been said elsewhere in this forum.

Upvotes: 3

Related Questions