Scheduling a Amazon SageMaker Notebook Instance example.

analysis.py

# Packages
import pandas as pd
import matplotlib.pyplot as plt
import boto3

# Default parameters
bucket = 'rtkilian-writing'
image_name = 'covid_cumulative_aus_state.png'

# Read
df = pd.read_csv('https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv')

# Data processing
df_aus = df.loc[df['Country/Region']=='Australia'] # filter for Australia
df_aus = df_aus.drop(['Country/Region', 'Lat', 'Long'], axis=1) # remove columns
df_aus = df_aus.transpose() # Transpose
df_aus = df_aus.rename(columns=df_aus.iloc[0]).drop(df_aus.index[0]) # Put the top row as the heading
df_aus.index = pd.to_datetime(df_aus.index) # Convert the index to datetime format
df_aus = df_aus[-365:] # last 365 days of data

# Visualise
ax = df_aus.plot(figsize=(12,6), title='Cumulative COVID-19 Confirmed Cases in Australia by State in the Past Year')
plt.show()

# Export locally
fig = ax.get_figure()
fig.savefig(image_name) # save the plot to local volume

# Export to S3
key='sagemaker-schedule/'+image_name
img_data = open(image_name, 'rb')
s3 = boto3.resource('s3')
s3.Bucket(bucket).put_object(Key=key, Body=img_data, 
                                 ContentType='image/png')