How to properly plot a bar chart in a Pandas dataframe? The x-values keep showing up as the index

Question

I have a small dataframe with student_id, exam_1, exam_2, exam_3, exam_4, and exam_5 as columns. There are 5 students as well for the rows. What I'd like to do is plot a bar graph showing the exam grades of one student aka one specific row, and ultimately doing it for each or a specific student from user input.

For now, though, I'm stuck on how to plot a bar graph for just one specific student.

import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({'student_id': [83838, 16373, 93538, 29383, 58585],
                   'exam_1': [80, 95, 90, 75, 50],
                   'exam_2': [60, 92, 88, 85, 40],
                   'exam_3': [70, 55, 75, 45, 60],
                   'exam_4': [55, 95, 45, 80, 55],
                   'exam_5': [91, 35, 92, 90, 75]})

print(df)

Which produces this as output:

  student_id  exam_1  exam_2  exam_3  exam_4  exam_5
0       83838      80      60      70      55      91
1       16373      95      92      55      95      35
2       93538      90      88      75      45      92
3       29383      75      85      45      80      90
4       58585      50      40      60      55      75

Adding this code below will allow me to select just one specific student ID aka row:

df = df.loc[df['student_id'] == 29383]
print(df)

   student_id  exam_1  exam_2  exam_3  exam_4  exam_5
3       29383      75      85      45      80      90

From here is where I'd like to plot this particular student's exams in a bar plot.

I tried the code below but it doesn't display it how I'd like. It seems that the index of this particular student is being used for the tick on the x-axis, if you can see the image. It will show '3' with some bar plots around it.

exam_plots_for_29383 = df.plot.bar()
plt.show()

Which will output this bar plot: Dataframe with bar plot. Looks weird.

I tried to transpose the dataframe, which kind of gets me to what I want. I used this code below:

df = df.T
exam_plots_for_29383_T = df.plot.bar()
plt.show()

But I end up with this as a graph: Transpose of dataframe with bar plot. Looks weird still.

I'm a bit stuck. I know there's a logical way of properly plotting a bar plot from the dataframe, I just can't for the life of me figure it out.

I'd like the bar plot to have:

• Exams 1 through 5 show up on the x-axis.
• Their values on the y-axis.
• Each exam bar in separate color.
• The legend showing the colors.

I think the last two options are done automatically. It's just the first two that are breaking my brain. I appreciate any help or tips.

Here's the code in full in case anyone would like to see it without it being split like above.

import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({'student_id': [83838, 16373, 93538, 29383, 58585],
                   'exam_1': [80, 95, 90, 75, 50],
                   'exam_2': [60, 92, 88, 85, 40],
                   'exam_3': [70, 55, 75, 45, 60],
                   'exam_4': [55, 95, 45, 80, 55],
                   'exam_5': [91, 35, 92, 90, 75]})

print(df)

df = df.loc[df['student_id'] == 29383]
print(df)

exam_plots_for_29383 = df.plot.bar()
plt.show()

df = df.T
exam_plots_for_29383_T = df.plot.bar()
plt.show()  

Answer 1

You are very close. The issue is that your numeric-like student ID is messing up all of the plots (which is why ID 29383 is giving you a bar close to 30,000 in all of your graphs).

Set the 'student_id' to the index so that it doesn't get plotted and now you can plot each student separately slicing the index with .loc[student_id], or if you plot the entire DataFrame it will color each different student.

df = df.set_index('student_id')
df.loc[29383].plot(kind='bar', figsize=(4,3), rot=30)

Knowing there are 5 exams you can give each its own color if you really want. Use a categorical color palette (tab10). (This also only works with Series.plot)

from matplotlib import cm
df.loc[29383].plot(kind='bar', figsize=(4,3), rot=30, color=cm.tab10.colors[0:5])