How to make a layered bar chart using matplotlib

Question

For this question, I was provided the following information.

Data in code form:

order_data = {'Alice': {5: 'chocolate'},
              'Bob': {9: 'vanilla'},
              'Clair': {7: 'strawberry'},
              'Drake': {10: 'chocolate' },
              'Emma': {82: 'vanilla'},
              'Alice': {70: 'strawberry'},
              'Emma': {42: 'chocolate'},
              'Ginger': {64: 'strawberry'} }

I was asked to make a bar graph detailing this data. The bar graph and the code used to make it using Altair is provided below.

import altair

data = altair.Data(customer=['Alice', 'Bob', 'Claire', 'Drake', 'Emma','Alice', 'Emma', 'Ginger'],
    cakes=[5,9,7,10,82,70,42,64],
    flavor=['chocolate', 'vanilla', 'strawberry','chocolate','vanilla','strawberry','chocolate','strawberry'])
chart = altair.Chart(data)
mark = chart.mark_bar()
enc = mark.encode(x='customer:N',y='cakes',color='flavor:N')
enc.display()

Graph:

My question is: What is the best way to go about constructing this graph using matplotlib?

I know this isn't an unusual graph per say but it is unusual in the sense that I have not found any replications of this kind of graph. Thank you!

r-beginners · Accepted Answer · 2020-10-12 04:17:31Z

2

It has already been answered, but you can also graph it in pandas.plot.

import pandas as pd
data = pd.DataFrame({'customer':['Alice', 'Bob', 'Claire', 'Drake', 'Emma','Alice', 'Emma', 'Ginger'],
    'cakes':[5,9,7,10,82,70,42,64],
    'flavor':['chocolate', 'vanilla', 'strawberry','chocolate','vanilla','strawberry','chocolate','strawberry']})
df = pd.DataFrame(data)
df = df.pivot(index='customer',columns='flavor', values='cakes').fillna(0)
df.plot(kind='bar', stacked=True)

answered Oct 12, 2020 at 4:17

r-beginners

35.6k3 gold badges20 silver badges41 bronze badges

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

Comments

Léonard · Accepted Answer · 2020-10-12 03:29:24Z

1

Here is a reproduction of the Altair graph with Matplotlib. Note that I had to modify the order_data dictionary because a dict cannot be defined with multiple keys at once (so I had to group the dictionary by key values). Also note that some optionally styling statements are included to also mimic the style of Altair.

The trick is to use the bottom keyword argument of the ax.bar function. The following image is obtained from the code below.

import matplotlib.pyplot as plt

# data
order_data = {
    "Alice": {"chocolate": 5, "strawberry": 70},
    "Bob": {"vanilla": 9},
    "Clair": {"strawberry": 7},
    "Drake": {"chocolate": 10},
    "Emma": {"chocolate": 42, "vanilla": 82},
    "Ginger": {"strawberry": 64},
}

# init figure
fig, ax = plt.subplots(1, figsize=(2.5, 4))
colors = {"chocolate": "C0", "strawberry": "C1", "vanilla": "C2"}

# show a bar for each person
for person_id, (name, orders) in enumerate(order_data.items()):
    quantities = 0
    for order_id, (order, quantity) in enumerate(orders.items()):
        ax.bar(person_id, quantity, bottom=quantities, color=colors[order])
        quantities += quantity

# add legend
ax.legend([color for color in colors], bbox_to_anchor=(2.0, 1.0))

# remove top/right axes for style match
ax.spines["top"].set_visible(False)
ax.spines["right"].set_visible(False)
ax.grid(axis="y", zorder=-1)

# ticks
ax.set_xticks(range(len(order_data)))
ax.set_xticklabels([name for name in order_data], rotation="vertical")

answered Oct 12, 2020 at 3:29

Léonard

2,6601 gold badge23 silver badges40 bronze badges

2 Comments

infomaniac Over a year ago

Is there anyway separate entries with the same customer may be placed on the same bar in the chart? For example say I had Emma purchase x amount of chocolate cakes, Alice purchases x amount of cakes, and then Emma comes back and purchases x amount of vanilla cakes. Is there anyway to have Emma's second purchase automatically be added to her existing data point on the chart without having to reorganize order_data so that her second order is included as a key value of her already existing location in the data? Thanks!

Léonard Over a year ago

I think what you are looking for is group dictionary by keys with a defaultdict so you can create a new plot with the newly added data. If you want to do it in a real update (animated) way, you should have a look at creating animations with Matplotlib.

Collectives™ on Stack Overflow

How to make a layered bar chart using matplotlib

2 Answers 2

Comments

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related