[FIXED] How to combine two time series plots using python seaborn?

Issue

I have a training and testing time series dataset that I would like to combine togther, to show how well the forecast did predicting the testing dataset.

Here is the toy code to reproduce the data:

import pandas as pd
import seaborn as sns

train_date = ['2017-01-01T00:00:00.000000000', '2017-02-01T00:00:00.000000000',
       '2017-03-01T00:00:00.000000000', '2017-04-01T00:00:00.000000000',
       '2017-05-01T00:00:00.000000000', '2017-06-01T00:00:00.000000000',
       '2017-07-01T00:00:00.000000000', '2017-08-01T00:00:00.000000000',
       '2017-09-01T00:00:00.000000000', '2017-10-01T00:00:00.000000000',
       '2017-11-01T00:00:00.000000000', '2017-12-01T00:00:00.000000000',
       '2018-01-01T00:00:00.000000000', '2018-02-01T00:00:00.000000000',
       '2018-03-01T00:00:00.000000000', '2018-04-01T00:00:00.000000000',
       '2018-05-01T00:00:00.000000000', '2018-06-01T00:00:00.000000000',
       '2018-07-01T00:00:00.000000000', '2018-08-01T00:00:00.000000000',
       '2018-09-01T00:00:00.000000000', '2018-10-01T00:00:00.000000000',
       '2018-11-01T00:00:00.000000000', '2018-12-01T00:00:00.000000000',
       '2019-01-01T00:00:00.000000000', '2019-02-01T00:00:00.000000000',
       '2019-03-01T00:00:00.000000000', '2019-04-01T00:00:00.000000000',
       '2019-05-01T00:00:00.000000000', '2019-06-01T00:00:00.000000000',
       '2019-07-01T00:00:00.000000000', '2019-08-01T00:00:00.000000000',
       '2019-09-01T00:00:00.000000000', '2019-10-01T00:00:00.000000000',
       '2019-11-01T00:00:00.000000000', '2019-12-01T00:00:00.000000000',
       '2020-01-01T00:00:00.000000000', '2020-02-01T00:00:00.000000000',
       '2020-03-01T00:00:00.000000000', '2020-04-01T00:00:00.000000000',
       '2020-05-01T00:00:00.000000000', '2020-06-01T00:00:00.000000000',
       '2020-07-01T00:00:00.000000000', '2020-08-01T00:00:00.000000000',
       '2020-09-01T00:00:00.000000000', '2020-10-01T00:00:00.000000000',
       '2020-11-01T00:00:00.000000000', '2020-12-01T00:00:00.000000000']
test_date = ['2021-01-01T00:00:00.000000000', '2021-02-01T00:00:00.000000000',
       '2021-03-01T00:00:00.000000000', '2021-04-01T00:00:00.000000000',
       '2021-05-01T00:00:00.000000000', '2021-06-01T00:00:00.000000000',
       '2021-07-01T00:00:00.000000000', '2021-08-01T00:00:00.000000000',
       '2021-09-01T00:00:00.000000000', '2021-10-01T00:00:00.000000000',
       '2021-11-01T00:00:00.000000000', '2021-12-01T00:00:00.000000000']
train_eaches = [1915.0, 1597.0, 1533.0, 1601.0, 1585.0, 1675.0, 1760.0, 1910.0, 1886.0, 1496.0, 1545.0, 1538.0, 1565.0, 1350.0,1686.0, 1535.0, 1629.0, 1589.0, 1605.0, 1560.0, 1353.0,1366.0, 1246.0, 1423.0, 1579.0, 1368.0, 1727.0, 1687.0, 1872.0, 1824.0, 2161.0, 1065.0, 727.0, 1567.0, 1509.0, 1687.0, 1647.0,1476.0, 1231.0, 1165.0, 1341.0, 1425.0, 1502.0, 1450.0, 1497.0, 1259.0, 1207.0, 1132.0]
test_eaches = [1252.0, 1038.0, 1184.0, 1200.0, 1219.0, 1339.0, 1504.0, 2652.0, 1724.0, 1029.0,
 711.0, 1530.0]
test_predictions = [1914.7225, 1490.4715, 1317.4765, 1341.263375, 1459.5875, 1534.2375, 1467.208875, 1306.2145, 1171.652625, 1120.641, 1138.912, 1171.914125]
test_credibility_down = [1805.   , 1303.   , 1017.   ,  915.975,  870.975,  797.   ,
        657.   ,  507.   ,  392.   ,  320.   ,  272.   ,  235.   ]
test_credibility_up = [2029.025, 1702.   , 1681.025, 1908.   , 2329.05 , 2695.025,
       2867.075, 2835.   , 2815.075, 2949.   , 3278.025, 3679.   ]

train_df = pd.DataFrame.from_dict({'date':train_date, 'eaches':train_eaches})
test_df = pd.DataFrame.from_dict({'date':test_date, 'eaches':test_eaches, '2.5% Credibilty':test_credibility_down,
                                 '97.5% Credibility':test_credibility_up})

Here are the two plots (train and test) and code that produces those plots:

fig = plt.figure(figsize=(15,4))
c=sns.scatterplot(x =train_df['date'], y = train_df['eaches'], label = 'Train Eaches',
               color = 'black')

fig = plt.figure(figsize=(15,4))

a=sns.lineplot(x =test_df['date'], y = test_df['predictions'], label = 'Posterior Prediction', color = 'red')
b=sns.lineplot(x =test_df['date'], y = test_df['2.5% Credibilty'], label = 'Credibilty Interval',
               color = 'skyblue', alpha=.3)
c=sns.lineplot(x =test_df['date'], y = test_df['97.5% Credibility'], label = 'Credibilty Interval',
               color = 'skyblue', alpha=.3)
line = c.get_lines()
plt.fill_between(line[0].get_xdata(), line[1].get_ydata(), line[2].get_ydata(), color='skyblue', alpha=.3)

sns.scatterplot(x =test_df['date'], y = test_df['eaches'], label = 'True Value', color='black')
plt.legend()

I would like to basically add the two x axis as a continuation and maybe add a vertical line to the start of the test period.

Solution

Put them on the same axes and use axvline to mark the prediction start. Also, you can fix the overlapping dates on the x-axis by casting the date columns as proper datetimes (train_df["date"] = pd.to_datetime(train_df.date)).

import matplotlib.pyplot as plt
import seaborn as sns
fig, ax = plt.subplots(1, 1, figsize=(15,4))
c_train= sns.scatterplot(x =train_df['date'], y = train_df['eaches'], label = 'Train Eaches',
                    color = 'black', ax=ax)
a = sns.lineplot(x =test_df['date'], y = test_df['predictions'], label = 'Posterior Prediction', color = 'red', ax=ax)
b = sns.lineplot(x =test_df['date'], y = test_df['2.5% Credibilty'], label = 'Credibilty Interval',
               color = 'skyblue', alpha=.3, ax=ax)
c = sns.lineplot(x =test_df['date'], y = test_df['97.5% Credibility'], label = 'Credibilty Interval',
               color = 'skyblue', alpha=.3)
line = c.get_lines()
ax.fill_between(line[0].get_xdata(), line[1].get_ydata(), line[2].get_ydata(), color='skyblue', alpha=.3)

sns.scatterplot(x =test_df['date'], y = test_df['eaches'], label = 'True Value', color='black', ax=ax)
ax.legend()

ax.axvline(test_df['date'][0])

Answered By - mcsoini