How to make the same crosstab line chart by using seaborn or matplotlib?-CodePudding

I have three columns. But I'm interested in plotting two columns with respect to each other. The datatype of the columns are datetime and object. The datetime column name is started_at and object column name is member_casual. I want started_at(hourly) on my x-axis and member_causal(count) on y-axis. I use this code

    my DaraFrame(df) looks like this
    
rideable_type        started_at             member_casual
electric_bike    2021-06-13 14:31:28            casual
classic_bike     2021-06-04 11:18:02            member
classic_bike     2021-06-04 09:49:35            casual
calssic_bike     2021-06-04 09:55:34            casual
electric_bike    2021-06-04 14:05:51            casual    
classic_bike     2021-06-04 14:09:59            member
electric_bike    2021-06-03 19:32:01            casual
electric_bike    2021-06-10 16:30:10            member
   ...                ...                         ...
   ...                ...                         ...

this code gives me line plot of counting member_casual(y-axis) with respect to hourly time(x-axis).

pd.crosstab(df.started_at.dt.hour,df.member_casual).plot(kind='line', ax = ax)

Can anyone tell me how to draw the same plot using seaborn or matplotlib ?

I,m begineer so this is the best I can do.

try_1

 plt.plot(df.started_at.dt.hour, df.member_casual)

try_2

sns.lineplot(df.started_at.dt.hour, hue = df.member_casual)

ofcourse both lines of code gives me an error.

CodePudding user response：

You first need to count the hourly occurrences, for instance the same way as you already did:

ct = pd.crosstab(df.started_at.dt.hour, df.member_casual)

matplotlib step by step:

plt.plot(ct.index, ct.casual, label='casual')
plt.plot(ct.index, ct.member, label='member')
plt.legend()

matplotlib short form:

plt.plot(ct)
plt.legend(ct.columns)

seaborn:

sns.lineplot(data=ct)