Adding more bars to plot

Question

I have an excel file looking like this:

| Dept.               | Q1 Budget | Q1 Actual | Q2 Budget | Q2 Actual | Q3 Budget | Q3 Actual | Q4 Budget | Q4 Actual |
|---------------------|-----------|-----------|-----------|-----------|-----------|-----------|-----------|-----------|
| Distribution        | 470,000   | 476,300   | 476,300   | 482,000   | 400,000   | 470,000   | 610,000   | 408,000   |
| Facilities          | 676,300   | 674,000   | 800,500   | 847,000   | 763,000   | 714,000   | 763,000   | 770,000   |
| Human Resources     | 363,000   | 432,000   | 463,000   | 442,000   | 463,000   | 440,000   | 443,000   | 467,000   |
| Information Systems | 763,000   | 726,300   | 863,000   | 870,000   | 876,300   | 776,000   | 700,000   | 740,000   |

I want to plot the quarters next to each other with the department as the label, with actual and budget as grouped bars, and the plot i would want to end up with would look something like this:

This is my code thus far:

df = pd.read_excel(file)

cols = df['Dept.']
df = df.T
df.columns = cols
df = df.tail(-1)

x = np.arange(4)
width = 0.30

plt.bar(x-0.15, df.loc['Q1 Actual'], width, color='cornflowerblue')
plt.bar(x+0.15, df.loc['Q1 Budget'], width, color='cornflowerblue', alpha=0.2)
plt.xticks(x, cols, rotation=45)

And this gives me the plot i want for the 1st quarter, looking like this:

How would i go about to extend my plot, to include every quarter?

Answer 1

Your data is in wide format, you would find it easier to do this if you melt your data into long format. Additionally the seaborn plotting library makes using a categorical variable for coloring the bars very simple.

import seaborn as sns
import pandas as pd

# Make plot bigger
sns.set(rc={'figure.figsize':(16,9)})

df = pd.DataFrame({'Dept.':['Q1 Actual','Q1 Budget', 'Q2 Actual', 'Q2 Budget'],
                  'Distribution':[476300, 470000, 482000, 476300],
                  'Facilities':[674000, 676300, 847000, 800500],
                  'Human Resources': [432000, 363000, 442000, 463000],
                  'Information Systems': [726300, 763000, 870000, 863000]})

df = df.melt(id_vars='Dept.', var_name='Function', value_name='Expenditure')

p = df.loc[df['Dept.'].str.startswith('Q1')]
sns.barplot(data=p, x='Function',y='Expenditure', hue='Dept.');

Alternatively you could plot them all at once with:

df['Quarter'] = df['Dept.'].str.split(' ').str[0]
g = sns.FacetGrid(df, col='Quarter', height=10)
g.map_dataframe(sns.barplot, x='Function', y='Expenditure', hue='Dept.')