In this post, I will describe how to position the intersection of x and y axis at a specific point using Matplotlib.
In order to re-position x and y axis, we need to understand an important concept in Matplotlib — spines. According to Matplotlib documentation:
Spines are the lines connecting the axis tick marks and noting the boundaries of the data area. They can be placed at arbitrary positions. See function:set_position for more information.
In normal plot, we can clearly see four spines around the plot. Like the following plot:
Two of the 4 spines, i.e., bottom and left spines are used as x and y axis. Another two spines are used to indicate data boundary.
Move the spines
In order to move a specific spine, we use the
method of spines. This method accepts a 2-tuple as its parameter. The first
element of the tuple is used to indicate the desired cooridinate system1,
data. If you use
axes, the value for the second element is
in the range $[0.0, 1.0]$. If you choose
data, the value for the second
element is the range of actual data range.
we can use
ax.spines to get the four spines. The returned object will be a
dict, whose keys are
right. Then we use one of these
keys to access a specific spine and customize it.
Postion spines in data coordinate
Let us take a concrete example. Suppose we want to put the intersection of x and y axis at data coordinate $(0, 0)$, we can use the following snippet:
ax.spines['left'].set_position(('data', 0)) ax.spines['bottom'].set_position(('data', 0))
The produced image will be:
The image looks a bit ugly since top and right spines now seem redundant. How do we remove them? It is easy:
Now the image will be prettier:
You can move the spines to whatever position you want, see more example below:
Postion spines in axes coordinate
We can also position spines using the axes coordinate system. For example, if you want to place the x and y axis in the top right of the axes (i.e., in the axes coordinate $(1, 1)$), use the following script:
ax.spines['left'].set_position(('axes', 1)) ax.spines['bottom'].set_position(('axes', 1))
The produced plot is
License CC BY-NC-ND 4.0