CODEWITHSUNDEEP
pythonpandasmatplotlib
Nikko
Nikko

Reputation: 1572

Subplot a groupby then Plot by groupby

Hi I am trying to plot a categorical data which is description:

data_df = pd.DataFrame({'Date': ['2018-09-14 00:00:22',
                                '2018-09-14 00:01:46',
                                '2018-09-14 00:01:56',
                                '2018-09-14 00:01:57',
                                '2018-09-14 00:01:58',
                                '2018-09-14 00:02:05'],
                        'userID': [33, 33, 33, 20, 20, 20],
                        'device': ['LIGHT', 'LIGHT', 'FAN', 'LIGHT', 'FAN', 'FAN'],
                        'description': ['ON', 'DIM', 'ON', 'ON', 'ON', 'OFF']})

data_df

What I want to is to plot them by grouping the 'device' but before that I want to subplot by rows the 'userID' so that they are separated each by 'userID' and each line plots are grouped based on the 'device' name.

I also tried grouping by and plot but it say my description is not numeric since it is categorical. Date is the x-axis, and description as y-axis.

Upvotes: 1

Views: 79

Answers (1)

Vishnu Kunchur
Vishnu Kunchur

Reputation: 1726

It looks like you might be trying to plot the description of userID and device groups over time. Obviously, description needs to be converted to a numeric variable, so I took the liberty of coding ON as 1, DIM as 0.5, and OFF as 0. The following code should solve your problem.

data_df['Date'] = pd.to_datetime(data_df['Date'])

def desc_num(x):
    if x == 'ON':
        return 1
    elif x == 'DIM':
        return 0.5
    else:
        return 0

data_df['desc_num'] = data_df['description'].apply(desc_num)    

## Creating groups of `userID` and `device`
groups = data_df.groupby(['userID', 'device'])

for g in groups:
    plt.plot(g[1]['Date'], g[1]['desc_num'])
    plt.xlabel('Time')
    plt.ylabel('description Status')
    plt.title('Time Series of userID: {0}, for device: {1}'.format(g[1]['userID'][0], g[1]['device'][0]))
    plt.show()
    plt.close()

Output (you should see such a plot for every subgroup):

enter image description here

Upvotes: 1

Related Questions