1

My current dataframe looks like this:

                     a   b    c    d    e     
 in_1   |   in_2  |
--------|-----------------------------------
 car    | bmw        2   4    5    34   46
        | merc       23  4   55    64   21 
        | range      453 32   2    56   21
        | lambo      4   6    2    5    12 
        | ferrari    12  46   34   23   642
fastfood| burger     123 34   213  23   234
        | kfc        123 34   235  123  24
        | tacoBell   213 432  124   12  1

I am trying to plot a subplot for each 'in_1' in which the x-axis is the column names (a, b, c, d, e), while the y-axis is the counts (the numbers in the cells).

So the first subplot would have the title "car". The x-axis would have 'a','b', 'c', 'd', 'e'. The y-axis will have the counts for each of 'bmw', 'merc', 'range', 'lambo', 'ferrari'.

The subplots can be bar or line plots and the values of in_2 can be represented in the form of a legend.

Improve this question
0

2 Answers 2

Reset to default
1

So I guess you could do something like this:

import numpy as np
import matplotlib.pyplot as plt

ng = 5 #number of groups

bmw = ..
merc = ..
..

fig, ax = plt.subplots()

index = np.arrange(ng)
bar_width = 0.2

fbmw = ax.bar(index, bmw, bar_width, color='r', label='BMW')
fmerc = ax.bar(index + bar_width, merc, color='b', label='MERC')
...
#don't forget to increase bar_width everytime

ax.set_xlabel('Cars')
ax.set_ylabel('Whatever this is')
ax.set_xticks(index + bar_width/2)
ax.set_xticklabels(('a', 'b', 'c', 'd', 'e'))
ax.legend()

fig.tight_layout()

Since I don't know what the numbers and columns a,b,c,d,e are, I left these labels empty. Also I thought you already have the dataframes for bmw, merc etc, so I didn't import them. Hope this helps!

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

1 Comment

I actually found something exteremely similar to what I have written. I guess thats the page I first learned about this stuff.matplotlib.org/gallery/statistics/barchart_demo.html
1

You can use a simple loop to pick up all your columns and assign them to an axis. It will also create a subplots with the number of rows determined by the unique values in in_1.

Please note that it assumes you have a mutli index df with in_1 and in_2 as index values.

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

n = len(df.columns)
rows = df.index.get_level_values('in_1').unique()
index = np.arange(n)

fig, axs = plt.subplots(len(rows),1,figsize=(15,10))
width = 0.2

colors=["#e08283", "#52b3d9", "#fde3a7", "#3fc380"]

for i in range(len(rows)):
    intdf = df[df.index.get_level_values('in_1') == rows[i]]
    offset = 0
    for j in range(len(intdf.index.get_level_values('in_2'))):
        v = intdf.iloc[j,].values
        axs[i].bar(index + offset, v, width, 
                   label=intdf.index.get_level_values('in_2')[j], color=colors[j])

        offset += width

    axs[i].set_xlabel(rows[i])
    axs[i].set_xticks(index + width)
    axs[i].set_xticklabels(tuple(df.columns))
    axs[i].legend(loc=2)


plt.show()

Here is the output.

enter image description here

As a reminder more information can be found here

Improve this answer

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.