CDS532 Programming for Data Science

Case Study Assignment

(Due: 11th December 2024, 23:59)

Introduction

Air pollution refers to the contamination of the atmosphere by the pollutants released into the air when industrial and commercial activities are conducted. These pollutants are harmful to human health and the environment. Three major sources of air pollutants in Hong Kong are motor vehicles, marine vessels, and power plants. In this case study, we would like to investigate the air pollution problem in Hong Kong. To achieve our objective, we will write a python program that can automatically fetch data from the Hong Kong government’s data portal and visualize the data into charts.

Program Requirement

Write a program that prompts the user to select which year’s air quality health index (AQHI) dataset should be used for conducting air pollution data visualization. The program will then fetch the dataset that correspond to the year selected by the user from the data resource in JSON format in the data portal named "Past record of Air Quality Health Index (English Version)". The API for fetching data is shown as follows:

This API takes a JSON string which contains the parameter values for configuring which data should be fetched. The definitions of the parameters are listed below:

After the AQHI data has been fetched, the program reads the AQHI data in JSON format downloaded from the data portal and conducts data preprocessing. A preview of the annual AQHI data is shown below:

The annual AQHI data is a table where each row corresponds to 16 different AQHI values measured in 16 air quality monitoring stations in a particular hour in a particular day. The first column is the date when the AQHI value is

measured. The second column is the hour when the AQHI value is measured. If a row has a value "Daily Max" in the hour column, that row is recording the maximum AQHI values of every air quality monitoring station in a particular  day. Each of the remaining 16 columns represents the AQHI data collected by one of the 16 air quality monitoring    stations located in the following 16 locations: Central/Western, Eastern, Kwun Tong, Sham Shui Po, Kwai Chung,

Tsuen Wan, Tseung KwanO, Yuen Long, Tuen Mun, Tung Chung, Tai Po, Sha Tin, Tap Mun, Causeway Bay, Central, Mong Kok.

Some AQHI values in the annual AQHI data may contain two extra symbols such as "+" and "*". These two symbols provide additional contextual information. However, these symbols may interfere with our analysis and have to be removed. On the other hand, some AQHI values maybe missing (i.e. empty value). Missing values are handled by    filling with zeros.

After performing data preprocessing, the program then performs the necessary calculations to plot the following two kinds of charts to visualize air population in the monthly scale:

The first kind of chart is a bar chart which plots the monthly average AQHI of16 air quality monitoring stations in Hong Kong.

The second kind of chart is a line chart which plot the changing trend of the AQHI averaged across 16 air quality monitoring stations in Hong Kong.

Note: To keep things simple, the maximum AQHI value X measured at monitoring station S in date D will be considered as the overall AQHI value of the monitoring station S at date D.

Sample Input

Sample output

Assumptions

You may assume that every input of the program is valid in format.

Submission

Students should submit their source code as (1) a single Jupiter Notebook file (i.e. .ipynb file) OR (2) a zip file that contains standalone Python script. files (i.e. .py files) for answering the programing questions to the submission box on the Moodleelearning platform. on or before 11th  December 2024, 23:59. Students are expected to name their file submission in the name of _case_study.ipynb OR _case_study.zip and their source code should follow the following format: