Python help

READ THIS FIRST

This assignment can be done in pairs of 2 people, or individually, and is divided into two parts.
First, you will work through a simple Jupyter Notebook intended to give you practice with pandas—think of this mainly as an exercise that will not receive feedback.
In part 2, the main focus, you will work with a large non-harmonized dataset (i.e., a dataset split across many Excel files). Your task is to harmonize the dataset (i.e., “combine it”) and transform it (clean it up and decide how to structure the data). In the second part you will also need to use pandas—so the knowledge from part 1 will be useful!

Download the Take-Home Exam

https://classroom.github.com/a/qnAGtjv7

Create a team, or join a team if you’ve chosen to work with a partner.

Accept the invitation.

Click the link to your repository.

Clone the repository by clicking “Code” → copy the link → git clone <link> in your terminal.

Open the project in VS Code via File → Open Folder, and you’re ready to start coding.

Note: This assumes you understand how Git works. Once you’ve cloned the repository you’re expected to make commits. In summary, you should run:

git add -A
git commit -m "a SHORT message about what I've done in this commit"
git push


If you’re working with a partner, either of you can make commits—it doesn’t matter.
It’s recommended to sit together; Git collaboration is not the focus of this lab, though it’s encouraged.

Assignment Description

Form groups of 2 people.
It’s OK to work alone if you ask for permission, but working in pairs is preferred.

Follow the instructions in part 1 and part 2.
You must record a 10–15 minute video where both of you are visible on camera and talk about your code.
More details in the section “Record a Video” further down.

Part 1 – Pandas

Read the instructions in the provided notebook—you are expected to complete it together (splitting tasks such as “you do 1 & 2, I do 4 & 5” is strictly forbidden).
Instead, always work together on Discord or in person.
Take turns writing the code.
Tip: use the Live Share extension in VS Code so you can code in the same editor at the same time.

Part 2 – Data Transformation and Harmonization

Create a Jupyter Notebook where you analyze and transform the data.
The goal is that, in the video, you can clearly explain how you analyzed and transformed the dataset, and that others can follow the decisions you made when cleaning/wrangling the data.

Transforming data (the T in ETL) can also mean enrichment—think about adding new columns, either derived from existing columns or bringing extra value (for example: for each application it might be useful to add a link to the decision, an address for each school, whether a school belongs to a specific organization, or if there are sub-categories beyond the main categories).

Dataset

https://www.myh.se/yrkeshogskolan/resultat-ansokningsomgangar/resultat-for-program

Choose a time range of at least 2–3 years—challenge yourself!

Note:
You are expected to use AI to help with the task, especially for transforming the data and writing pandas queries.
Just make sure you understand what the code does.

Grading: U, G, VG
G (Pass)

Complete the pandas notebook.

Show understanding of transformation and harmonization of the YH dataset.

Present naturally in the video, talking through the notebook and alternating with showing the raw data



Recording the Video (You will just prepare me for this part)

Here’s a simple tutorial for recording with a facecam using OBS:
https://www.youtube.com/watch?v=uL8BwstqiqE&ab_channel=KevinStratvert

Upload the video to Google Drive and make it public—test the link in an incognito browser to ensure access.
Put the link in the file videolink.txt in your repository.
Do not send it via Discord.

Recording with Two Cameras

If you are two people, an easy way is to start a Discord video call and place that window on the right side of your screen so both of you are visible.
This can also be done in OBS using multiple sources, but the Discord method is simplest.

Steps:

Start a Discord call with video.

Select your entire screen as a source in OBS.

Make sure OBS captures audio from all sources.

Use VS Code Live Share so the person not screen-sharing can control the code when needed.

It’s OK to do light editing, but don’t overdo it.

Ensure video quality is at least 720p, that both of you are visible and audible.

Make sure the video is publicly accessible—I shouldn’t need to request access.

Good luck!