[FIXED] how to get seperate line charts using loop in jupyter notebook on groupby values

Issue

I have a data frame of this format.

brand-var1-var2(date)-var3
A - 100 - 20/12 - 300
A - 110 - 12/12 - 132
B - 24 - 24/12 - 543
C - 235 - 3/12- 534

I want to print individual separate line chart in jupyter notebook for each brand such that

For brand A

x_axis = df.var2
y_axis = df.var1

then for Brand B

x_axis = df.var2
y_axis = df.var1

then for Brand C and so on...

I tried using Loop with matplotlib but in vain.

from pandas import *
import matplotlib.pyplot as plt
%matplotlib inline

ys = df['orders'], df['gmv']
x_ax = df['brand']

for y_ax in ys:
    ts = Series(y_ax,index=x_ax)
    ts.plot(kind='bar', figsize=(15,5))
    plt.show()

This shows error - ValueError: cannot reindex from a duplicate axis

Thanks

Solution

I'm not sure what type of plot you want. At the beginning of your question, you ask for a line plot but in the code you call kind='bar'. Since the dataframe looks like it contains times series data, I am going to assume you want a line plot.

I created a random dataframe with the following:

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

def make_random_date():    
    year = np.random.choice(range(2000, 2010))
    month = np.random.choice(range(1, 13))
    day = np.random.choice(range(1, 29))
    return datetime(year, month, day)

# Make a random dataframe
n_samples = 100
df = pd.DataFrame()
df['date'] = [make_random_date() for _ in range(n_samples)]
df['brand'] = [np.random.choice(['A','B','C']) for _ in range(n_samples)]
df['var1'] = np.random.randint(0, 100, size=(n_samples, 1))
df.head()

    date    brand   var1
0   2003-03-27  B   16
1   2009-06-24  C   7
2   2008-04-17  A   82
3   2004-02-20  C   9
4   2007-05-10  B   69

Then to create individual line plots for each brand, do the following:

for brand in ['A', 'B', 'C']:
    sub = df.query("brand == @brand")
    ts = pd.Series(data=list(sub['var1']), index=sub['date'])
    ts.plot()
    plt.title(f'Brand: {brand}')
    plt.show()

And this is the output:

Answered By - vivek