Matplotlib- Creating a table with line plots in cells?

Question

I have historical baseball data that I'm trying to visualize in a simple matplotlib plot Within 1 subplot I want to have a table that shows average statistics over the past year, a line chart for each statistic, and then a final score which is calculated independently.

I know matplotlib has a table function so it would be simple to create a 5x3 table, but is it possible to insert plots as a value in a table? If not, are there any suggestions on what I should do? I guess I could create several subplots, but formatting would be wacky and not very dynamic. Appreciate the help.

Using @TheImportanceofBeingErnest's code, I'm encountering a bug with matplotlib where I can't see the x axis when I use gridspec:

fig = plt.figure(figsize=(8.5,11))

gs_row1left = gridspec.GridSpec(1,1)
gs_row1right = gridspec.GridSpec(1,1)

summaryplot2subplot(fig, gs_row1left[0], data, col1, col2, finalsc)
ax.axis('off')

ax2 = fig.add_subplot(gs_row1right[0, 0])
df = pd.DataFrame({'year':['2001-01','2002-01','2003-01','2004-01','2005-01'], 'value':[100,200,300,400,500]})

barax = ax2.twinx()
df['value1']= df['value']*0.4

df['value2'] = df['value']*0.6# Let them be strings!



df.plot(x = ['year'], y = ['value'], kind = 'line', ax = ax2)

df.plot(x = ['year'], y= ['value1','value2'], kind = 'bar', ax = barax)



gs_row1left.update(left = 0.05, right = 0.48)
gs_row1right.update(left = 0.55, right = 0.98)
plt.show()

ImportanceOfBeingErnest · Accepted Answer · 2018-02-06 21:56:30Z

3

It is not possible to insert plots into a matplotlib table. However subplot grids allow to create a table-like behaviour.

import matplotlib.pyplot as plt
import numpy as np

data = np.random.rand(100,4)
col1 = ["WAR", "ERA", "IP", "WHIP", "Final\nScore"]
col2 = [0.23,1.60,0.28,0.02,0.38]
col2colors = ["red", "g", "r", "r", "r"]
finalsc = "D+"

fig, axes = plt.subplots(ncols=3, nrows=5, figsize=(4,2.6),
                         gridspec_kw={"width_ratios":[1,0.5,2]})
fig.subplots_adjust(0.05,0.05,0.95,0.95, wspace=0.05, hspace=0)

for ax in axes.flatten():
    ax.tick_params(labelbottom=0, labelleft=0, bottom=0, top=0, left=0, right=0)
    ax.ticklabel_format(useOffset=False, style="plain")
    for _,s in ax.spines.items():
        s.set_visible(False)
border = fig.add_subplot(111)
border.tick_params(labelbottom=0, labelleft=0, bottom=0, top=0, left=0, right=0)
border.set_facecolor("None")

text_kw = dict(ha="center", va="bottom", size=13)
for i,ax in enumerate(axes[:,0]):
    ax.text(0.5, 0.05, col1[i], transform=ax.transAxes, **text_kw)
for i,ax in enumerate(axes[:,1]):
    ax.text(0.5, 0.05, "{:.2f}".format(col2[i]),transform=ax.transAxes, **text_kw)
    ax.set_facecolor(col2colors[i])
    ax.patch.set_color(col2colors[i])
axes[-1,-1].text(0.5, 0.05, finalsc,transform=axes[-1,-1].transAxes, **text_kw)

for i,ax in enumerate(axes[:-1,2]):
    ax.plot(data[:,i], color="green", linewidth=1)


plt.show()

To put several such plots into a figure you would approach this a bit differently and create a gridspec with several subgrids.

import matplotlib.pyplot as plt
from matplotlib import gridspec
import numpy as np


def summaryplot2subplot(fig, gs, data, col1, col2, finalsc):
    col2colors = ["g" if col2[i] > 1 else "r" for i in range(len(col2)) ]
    sgs = gridspec.GridSpecFromSubplotSpec(5,3, subplot_spec=gs, wspace=0.05, hspace=0,
                                           width_ratios=[0.9,0.7,2])
    axes = []
    for n in range(5):
        for m in range(3):
            axes.append(fig.add_subplot(sgs[n,m]))
    axes = np.array(axes).reshape(5,3)
    for ax in axes.flatten():
        ax.tick_params(labelbottom=0, labelleft=0, bottom=0, top=0, left=0, right=0)
        ax.ticklabel_format(useOffset=False, style="plain")
        for _,s in ax.spines.items():
            s.set_visible(False)
    border = fig.add_subplot(gs)
    border.tick_params(labelbottom=0, labelleft=0, bottom=0, top=0, left=0, right=0)
    border.set_facecolor("None")
    
    text_kw = dict(ha="center", va="bottom", size=11)
    for i,ax in enumerate(axes[:,0]):
        ax.text(0.5, 0.05, col1[i], transform=ax.transAxes, **text_kw)
    for i,ax in enumerate(axes[:,1]):
        ax.text(0.5, 0.05, "{:.2f}".format(col2[i]),transform=ax.transAxes, **text_kw)
        ax.set_facecolor(col2colors[i])
        ax.patch.set_color(col2colors[i])
    axes[-1,-1].text(0.5, 0.05, finalsc,transform=axes[-1,-1].transAxes, **text_kw)
    
    for i,ax in enumerate(axes[:-1,2]):
        ax.plot(data[:,i], color=col2colors[i], linewidth=1)


fig = plt.figure(figsize=(8,6))
gs = gridspec.GridSpec(2,2)


col1 = ["WAR", "ERA", "IP", "WHIP", "Final\nScore"]
finalsc = "D+"

for i in range(4):
    data = np.random.rand(100,4)
    col2 = np.random.rand(5)*2
    summaryplot2subplot(fig, gs[i], data, col1, col2, finalsc)

plt.show()

edited Feb 6, 2018 at 21:56

answered Dec 13, 2017 at 10:17

ImportanceOfBeingErnest

342k61 gold badges737 silver badges771 bronze badges

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

21 Comments

Dick Thompson Over a year ago

This is exactly what I need- thanks! I might be running an older version of matplotlib, but for some reason I had to insert a path for set_facecolor, ie ax.patch.set_facecolor(col2colors[i])

Dick Thompson Over a year ago

I'm trying to include this plot as a subplot of other plots in a 4x2 grid. Do you know if this is possible since we are already using subplots in this example? Appreciate the help

Dick Thompson Over a year ago

awesome- thanks so much! so if I just wanted to insert one of those into another plot (I basically just want to have one chart in a larger plot) how would I put all of that into one ax? would I have to use gridspec?

ImportanceOfBeingErnest Over a year ago

The larger plot would need to use gridspec. But it is only necessary for this table-like plot, all other subplots can potentially also be created differently.

Dick Thompson Over a year ago

so would it be possible to put all of this on one ax, where ax would be something like ax = fig.add_subplot(4,2,1)

|

Collectives™ on Stack Overflow

Matplotlib- Creating a table with line plots in cells?

1 Answer 1

21 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

21 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related