How to add a secondary x-axis with categorical features?

Question

After doing some research on this topic, I couldn't find an answer to this particular problem. I want to make a secondary x-axis but with categorical variables that repeat within intervals (but not repeating the same value over and over in the plot). A similar example of what I want can be seen in this picture (made with excel) :

The data:

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

data1 = {'Month': list(range(11,35))+list(range(34,42)),
         'Checkpoint': ['A','A','A','A','A','A','B','B','B','B','B','B','C','C','C','C','C','C','C','C','C','C','C','D','C','D','D','D','D','D','D','D'],
         'Litres':[216545.67,18034.45,25807.83,46136.23,68099.21,55436.35,56412.33,9347.52,3177.29,103.89,333.29,2355.41,
                     49063.72,113622.80,243639.97,303992.32,255471.55,267022.75,274952.92,619665.39,798969.54,1127476.60,
                     1563344.98,1051827.75,603167.32,1880605.49,1931002.19,
                     2970500.68,2362336.66,5311058.83,5071784.10,5325575.47]}

df = pd.DataFrame(data1)

By running the code above, we obtain the following dataframe

     Month   Checkpoint    Litres
0     11         A        216545.67
1     12         A        18034.45
2     13         A        25807.83
3     14         A        46136.23
4     15         A        68099.21
5     16         A        55436.35
6     17         B        56412.33
7     18         B        9347.52
8     19         B        3177.29
9     20         B        103.89
10    21         B        333.29
11    22         B        2355.41
12    23         C        49063.72
13    24         C        113622.80
14    25         C        243639.97
15    26         C        303992.32
16    27         C        255471.55
17    28         C        267022.75
18    29         C        274952.92
19    30         C        619665.39
20    31         C        798969.54
21    32         C        1127476.60
22    33         C        1563344.98
23    34         D        1051827.75
24    34         C        603167.32
25    35         D        1880605.49
26    36         D        1931002.19
27    37         D        2970500.68
28    38         D        2362336.66
29    39         D        5311058.83
30    40         D        5071784.10
31    41         D        5325575.47

I want to do a scatterplot (can be either with matplotlib or seaborn) of the data but with the 2nd x-axis (df['Checkpoint']).

plt.figure(figsize = (14,7))
plt.scatter(df['Month'], df['Litres'], s=30)

Answer 1

One possibility is to use the major ticks to put the month labels and the minor ticks to make separations. Whenever the checkpoint label changes, a longer tick needs to be drawn. The checkpoint label position will be just between two long ticks.

There seems to be one month with two different labels (month 34). It is not clear what has to happen then. In the code below a long major tick is drawn there.

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.ticker as ticker

data1 = {'Month': list(range(11, 35)) + list(range(34, 42)),
         'Checkpoint': ['A', 'A', 'A', 'A', 'A', 'A', 'B', 'B', 'B', 'B', 'B', 'B', 'C', 'C', 'C', 'C', 'C', 'C', 'C',
                        'C', 'C', 'C', 'C', 'D', 'C', 'D', 'D', 'D', 'D', 'D', 'D', 'D'],
         'Litres': [216545.67, 18034.45, 25807.83, 46136.23, 68099.21, 55436.35, 56412.33, 9347.52, 3177.29, 103.89,
                    333.29, 2355.41, 49063.72, 113622.80, 243639.97, 303992.32, 255471.55, 267022.75, 274952.92,
                    619665.39, 798969.54, 1127476.60, 1563344.98, 1051827.75, 603167.32, 1880605.49, 1931002.19,
                    2970500.68, 2362336.66, 5311058.83, 5071784.10, 5325575.47]}
df = pd.DataFrame(data1)

fig, ax = plt.subplots(figsize=(10, 5))

ax.scatter(df['Month'], df['Litres'], s=30, color='crimson')

ax.xaxis.set_major_locator(ticker.FixedLocator(0.5))
ax.xaxis.set_major_locator(ticker.MultipleLocator(1))
ax.xaxis.set_minor_locator(ticker.MultipleLocator(0.5))
ax.set_xlim(df['Month'].iloc[0] - 0.5, df['Month'].iloc[-1] + 0.5)

checkpoints = list(df['Checkpoint'])

long_minor_ticks = [df['Month'].iloc[0] - 1]  # these minor ticks need to be longer
long_major_ticks = []  # these major ticks need to be longer

for m1, m, c1, c in zip(df['Month'][1:], df['Month'], df['Checkpoint'][1:], df['Checkpoint']):
    if m == m1:
        long_major_ticks.append(m)
    elif c != c1:
        long_minor_ticks.append(m)
long_minor_ticks.append(df['Month'].iloc[-1])

ax.tick_params(which='minor', axis='x', pad=20) # put the minor tick labels at some distance

checkpoint_labels = []
for tick, month in zip(ax.xaxis.get_minor_ticks(), range(df['Month'].iloc[0]-1, 100)):
    l = 35 if month in long_minor_ticks and not month in long_major_ticks and not month+1 in long_major_ticks else 18
    tick.tick1line.set_markersize(l)
    checkpoint_labels.append('')

for tick, month in zip(ax.xaxis.get_major_ticks(), range(df['Month'].iloc[0]-1, 100)):
    l = 35 if month in long_major_ticks else 0
    tick.tick1line.set_markersize(l)

# set the checkpoint letters at the positions between the long minor ticks
for t0, t1 in zip(long_minor_ticks[:-1],long_minor_ticks[1:]):
    if t1 != t0 + 1:
        ind = (t1+t0) // 2 - long_minor_ticks[0]
        checkpoint_labels[ind] = df['Checkpoint'].iloc[ind]
ax.set_xticklabels(checkpoint_labels, minor=True)

fig.subplots_adjust(bottom=0.15) # we need space to show the large ticks
plt.show()