I have a table with three columns (Month_Year, SKU_ID, Actual_Demand)

I need to create a plot that showcases the movement in actual demand for each of my 10 SKU's.

The data is structured like this.

So far my method is to filter for each of the 10 SKU's and create a seperat plot.

# -*- coding: utf-8 -*-
"""
Created on Thu Sep  9 14:31:10 2021

@author: D996FFO
"""


'Importing Relevant Packages'

import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns



'Load in dataset with historical sales values'
ad = pd.read_excel (r'C:/Users/d996ffo/Model_Data.xlsx', sheet_name= 'Data_For_Python')


#SKU 1
sku_1 = ad.loc[ad['SKU_ID'] == 1]

'Converting the string format month year into a date format'
sku_1['Month_Year'] = pd.to_datetime(sku_1['Month_Year'])


'Set index equal to the date'
sku_1.index = sku_1['Month_Year']
del sku_1['Month_Year']
del sku_1['SKU_ID']

#SKU 2

sku_2 = ad.loc[ad['SKU_ID'] == 2]

'Converting the string format month year into a date format'
sku_2['Month_Year'] = pd.to_datetime(sku_2['Month_Year'])


'Set index equal to the date'
sku_2.index = sku_2['Month_Year']
del sku_2['Month_Year']
del sku_2['SKU_ID']


plt.plot(sku_1, color = 'blue', label = 'SKU 1')
plt.plot(sku_2, color = 'red', label = 'SKU 2')

sns.lineplot(data = sku_2.Actual_Demand)

But there has to be a better way than this?

I want to forecast based on each of the SKU later, when i have studied the data and it seems to me that i am not doing this smart.

CodePudding user response：

You can try using Seaborn and read up on the hue parameter.

For extra info, matplotlib is good for plotting 1-2variable while seaborn is great for more than 2 variables. In your case, you have 3 variables.

import seaborn as sns

#Assume ad is your DataFrame

sns.lineplot(data=ad, x="Month_Year", y="Actual_Demand", hue="SKU_ID")
plt.show()