如何在python中的函数之间传递/返回matplotlib对象

Question

I am trying to create a module that contains simple functions for creating plots with some common formatting already applied to them. Some of these functions would be applied to matplotlib objects that already exist and return other matplotlib objects to the main program.

This first segment of code is an example of how I currently generate plots and it works as is.

# Include relevant python libraries
from matplotlib import pyplot as plt

# Define plot formatting
axesSize = [0, 0, 1, 1]
axesStyle = ({'facecolor':(0.95, 0.95, 0.95)})

gridStyle = ({'color':'k',
              'linestyle':':',
              'linewidth':1})

xString = "Independent Variable"
xLabelStyle = ({'fontsize':18,
                'color':'r'})

# Create figure and axes objects with appropriate style
figureHandle = plt.figure()
axesHandle = figureHandle.add_axes(axesSize, **axesStyle)

axesHandle.grid(**gridStyle)
axesHandle.set_xlabel(xString, **xLabelStyle)

I want to create a function that combines the add_axes() command with the grid() and set_xlabel() commands. As a first attempt, ignoring all styling, I came up with the following function in my NTPlotTools.py module.

def CreateAxes(figureHandle, **kwargs):
    axesHandle = figureHandle.add_axes()
    return axesHandle

The script that calls the functions looks like:

# Include relevant python libraries
from matplotlib import pyplot as plt
from importlib.machinery import SourceFileLoader as fileLoad

# Include module with my functions
pathName = "/absolute/file/path/NTPlotTools.py"
moduleName = "NTPlotTools.py"
pt = fileLoad(moduleName, pathName).load_module()

# Define plot formatting
gridStyle = ({'color':'k',
              'linestyle':':',
              'linewidth':1})

# Create figure and axes objects with appropriate style
figureHandle = plt.figure()
axesHandle = pt.CreateAxes(figureHandle)

axesHandle.grid(**gridStyle)

However, I get the following error message when I run the main code:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-4-73802a54b21a> in <module>()
     17 axesHandle = pt.CreateAxes(figureHandle)
     18 
---> 19 axesHandle.grid(**gridStyle)

AttributeError: 'NoneType' object has no attribute 'grid'

This says to me that the axesHandle is not a matplotlib axes object, and by extension, the CreateAxes() function call did not return a matplotlib axes object. Is there a trick to passing matplotlib objects to/from functions?

Answer 1

You are almost there. The problem is with this line.

def CreateAxes(figureHandle, **kwargs)
    axesHandle = figureHandle.add_axes() # Here
    return axesHandle

From the source code the add_axes method will looks like the following

def add_axes(self, *args, **kwargs):
    if not len(args):
       return
    # rest of the code ...

So when you invoke figureHandle.add_axes() without any parameter both args and kwrags will be empty. From the source code if args is empty add_axes method returns None . Hence this None value gets assigned to axesHandle and when you try to invoke axesHandle.grid(**gridStyle) you will get

AttributeError: 'NoneType' object has no attribute 'grid'

Example

>>> def my_demo_fun(*args, **kwrags):
...     if not len(args):
...          return
...     return args
...
>>> print(my_demo_fun())
None
>>> print(my_demo_fun(1, 2))
(1, 2)

So re-write the function by passing arguments to the add_axes method.

def create_axes(figure_handle, **kwargs):
    axes_handle = figure_handle.add_axes(axes_size, **axes_style) 
    return axes_handle