Summary
Download the zip file from Canvas corresponding to your student id and unzip the contents into your local assignment folder and ensure your files are similar to Figure 1.
Assignment Sections
Data ETL –
Original Data Headings
car_reg purchase_date county
make
model type colour tax_band price
the car registration plate
the purchase date of the car
the county car was purchased & registered the car manufacturers name
the car model name
the type of car (e.g., saloon, hatchback etc.) the colour of the car
the tax band of the car
the purchase price of the car in Euros
Figure 1 – Example Assignment Folder & Files
Clean the dataset provided: data/R00000000_original.csv and save as data/R00000000_processed.csv replacing R00000000 with your CIT student number.
All code required to reproduce the data ETL process should be placed in the Python library file (at the bottom where indicated): lib/R00000000_util.py and able to be called from the Jupyter Notebook: R00000000_A1_Notebook.ipynb.
|
Column Name |
Column Description |
DATA8001 Assignment 1
Processed Data Headings & Expected Data Types
|
Column Name |
Column Description |
Data Type |
car_reg purchase_date year
month
county
make
model
type
colour tax_band price
Example
Cleaned car registration plate Cleaned purchase date of the car The year the car was purchased The month the car was purchased Cleaned county name
Cleaned car manufacturers name Cleaned car model name Cleaned car type
Cleaned colour of the car Cleaned tax band of the car
the purchase price of the car in Euros
String (uppercase) Datetime
Int
Int
String (uppercase) String (uppercase) String (uppercase) String (uppercase) String (uppercase) String (uppercase) Float
Data Visualisation – 10%
Load the processed dataset (data/R00000000_processed.csv) into the assignment notebook: R00000000_A1_Notebook.ipynb and answer the 5 questions including 1 (& only 1) visualisation of your choice that best answers each question. Show your workings in the Jupyter Notebook for each question.
Data Modelling – 10%
Create a Linear Regression model and any transformations required to give your model the best accuracy. Using the Python class provided in lib/R00000000_util.py, save the object to the model folder as: model/R00000000.pkl.
All code required to reproduce the modelling process should be placed in the Python library file: lib/R00000000_util.py and able to be called from the Jupyter Notebook: R00000000_A1_Notebook.ipynb.
The pickled model file should be loaded and called from the Jupyter Notebook and available to process unseen test data including any transformations required to ensure the model works. Note: the unseen test data will have the same headings & datatypes as your data/R00000000_processed.csv file.
Report & Questions (15%)
Write a max 2-page report outlining the steps taken to complete the assignment. Identify any areas you feel are worth mentioning during the ETL, visualisation of modelling steps including any insights developed.
Answer 2 exam type questions (max 300 words) each. Note – due to the “open-book” nature of this assignment, a clean, concise and well-thought-out answer of your “own” viewpoint is expected, this is not a “cut and paste” exercise!
DATA8001 Assignment 1
P