Matplotlib & Plotly Charts

Matplotlib

Using pictures.add(), it is easy to paste a Matplotlib plot as picture in Excel.

Getting started

The easiest sample boils down to:

import matplotlib.pyplot as plt
import xlwings as xw

fig = plt.figure()
plt.plot([1, 2, 3])

sheet = xw.Book().sheets[0]
sheet.pictures.add(fig, name='MyPlot', update=True)
_images/mpl_basic.png

Note

If you set update=True, you can resize and position the plot on Excel: subsequent calls to pictures.add() with the same name ('MyPlot') will update the picture without changing its position or size.

Note

If you get the following warning UserWarning: Starting a Matplotlib GUI outside of the main thread will likely fail., make sure to set the Matplotlib backend before using any Matplotlib/Pyplot code: import matplotlib as mpl and mpl.use("agg").

Full integration with Excel

Calling the above code with RunPython and binding it e.g. to a button is straightforward and works cross-platform.

However, on Windows you can make things feel even more integrated by setting up a UDF along the following lines:

@xw.func
def myplot(n, caller):
    fig = plt.figure()
    plt.plot(range(int(n)))
    caller.sheet.pictures.add(fig, name='MyPlot', update=True)
    return 'Plotted with n={}'.format(n)

If you import this function and call it from cell B2, then the plot gets automatically updated when cell B1 changes:

_images/mpl_udf.png

Properties

Size, position and other properties can either be set as arguments within pictures.add(), or by manipulating the picture object that is returned, see xlwings.Picture().

For example:

>>> sht = xw.Book().sheets[0]
>>> sht.pictures.add(fig, name='MyPlot', update=True,
                     left=sht.range('B5').left, top=sht.range('B5').top)

or:

>>> plot = sht.pictures.add(fig, name='MyPlot', update=True)
>>> plot.height /= 2
>>> plot.width /= 2

Getting a Matplotlib figure

Here are a few examples of how you get a matplotlib figure object:

  • via PyPlot interface:

    import matplotlib.pyplot as plt
fig = plt.figure()
plt.plot([1, 2, 3, 4, 5])

    or:

    import matplotlib.pyplot as plt
plt.plot([1, 2, 3, 4, 5])
fig = plt.gcf()

  • via object oriented interface:

    from matplotlib.figure import Figure
fig = Figure(figsize=(8, 6))
ax = fig.add_subplot(111)
ax.plot([1, 2, 3, 4, 5])

  • via Pandas:

    import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.rand(10, 4), columns=['a', 'b', 'c', 'd'])
ax = df.plot(kind='bar')
fig = ax.get_figure()

Note

When working with Google Sheets, you can use a maximum of 1 million pixels per picture. Total pixels is a function of figure size and dpi: (width in inches * dpi) * (height in inches * dpi). For example, fig = plt.figure(figsize=(6, 4)) with 200 dpi (default dpi when using pictures.add()) will result in (6 * 200) * (4 * 200) = 960,000 px. To change the dpi, provide export_options: pictures.add(fig, export_options={"bbox_inches": "tight", "dpi": 300}). Existing figure size can be checked via fig.get_size_inches(). pandas also accepts figsize like so: ax = df.plot(figsize=(3, 3)). Note that "bbox_inches": "tight" crops the image and therefore will reduce the number of pixels in a non-deterministic way. export_options will be passed to figure.figsave() when using Matplotlib and to figure.write_image() when using Plotly.

Plotly static charts

Prerequisites

In addition to plotly, you will need kaleido, psutil, and requests. The easiest way to get it is via pip:

$ pip install kaleido psutil requests

or conda:

$ conda install -c conda-forge python-kaleido psutil requests

See also: https://plotly.com/python/static-image-export/

How to use

It works the same as with Matplotlib, however, rendering a Plotly chart takes slightly longer. Here is a sample:

import xlwings as xw
import plotly.express as px

# Plotly chart
df = px.data.iris()
fig = px.scatter(df, x="sepal_width", y="sepal_length", color="species")

# Add it to Excel
wb = xw.Book()
wb.sheets[0].pictures.add(fig, name='IrisScatterPlot', update=True)
_images/plotly.png