dual axis chart with two secondary y axis

Question

I want to create a dual axis chart showing monthly user trips with the two secondary y axis showing average duration per month for each user type.

ds = {"duration":['1.02', '1.03', '1.08', '1.07', '2.02', '1,01'], 
      "start_time": ['2019-01-01 00:07:10.576', '2019-01-31 23:48:50.0920', '2019-01-01 00:11:03.441', '2019-01-31 20:58:33.8860', '2019-01-01 00:11:03.441', '2019-01-01 00:14:48.398'],
      "user": [0, 1, 1, 0, 1, 0]
     }
df = pd.DataFrame(ds)

Above is a sample of my dataframe with start_time in datetime format, duration in float, and user is a dummy variable where 0 represents "customer" and 1 represents "subscriber"

Seems your data has to be processing for monthly averages, can you add to question dtaframe for plot from your input data? — jezrael, Oct 24 '22 at 05:40

score 0 · Accepted Answer · answered Oct 24 '22 at 08:50

It makes more sense to have more than one month in your dataset so I've changed that

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

ds = {"duration": [1.02, 1.03, 1.08, 1.07, 2.02, 1.01],
      "start_time": ['2019-01-01 00:07:10.576', '2019-01-31 23:48:50.0920', '2019-01-01 00:11:03.441',
                     '2019-02-27 20:58:33.8860', '2019-02-01 00:11:03.441', '2019-02-01 00:14:48.398'],
      "user": [0, 1, 1, 0, 1, 0]
      }
df = pd.DataFrame(ds)

get the month of each start_time and replace 1 and 0 with subscriber and customer (just for display)

df["month"] = pd.to_datetime(df.start_time).dt.month
df.user = np.where(df.user, "subscriber", "customer")

Groupby and calculate necessary data

data = df.groupby(["user", "month"]).agg(trips=("user", "count"), mean_duration=("duration", "mean"))

Build the plot

fig, ax1 = plt.subplots()
ax2 = ax1.twinx()
data.trips.plot(kind='bar', color='red', ax=ax1, width=0.4, position=1)
data.mean_duration.plot(kind='bar', color='blue', ax=ax2, width=0.4, position=0)
plt.setp(ax1.get_xticklabels(), ha="right", rotation=45)  # rotate labels (looks nicer)
plt.tight_layout()
plt.show()  # show plot (might be unnecessary for you)

Let me know if you have questions or if I misunderstood anything

I love this, i'll try it and let you know if it works – Shedrack David Oct 25 '22 at 08:16 — Shedrack David, Oct 25 '22 at 08:16

dual axis chart with two secondary y axis

1 Answers1