Exercise 2: Collaboration - Research Team Simulation

Time: 25 minutes (Round 1: 10 min, Round 2: 10 min, Debrief: 5 min)

Pair up with someone near you. Decide who is Person A and who is Person B.

---

Setup (3 minutes)

Person A:

1. Go to GitHub → New Repository → name it collab-research
2. Clone it locally:

   git clone https://github.com/YOUR-USERNAME/collab-research.git
   cd collab-research
   

3. Copy the template files (provided by instructor) into this folder
4. Commit and push:

   git add .
   git commit -m "Add initial project files"
   git push
   

5. Go to repo Settings → Collaborators → Add people → add Person B's GitHub username

Person B:

1. Accept the invitation (check your email or GitHub notifications bell)
2. Clone the repo:

   git clone https://github.com/PERSON-A-USERNAME/collab-research.git
   cd collab-research
   

Both of you should now have the same files locally.

---

Round 1: Parallel Work - The Happy Path (10 minutes)

You'll each work on a different file so there's no conflict.

Person A: Improve the analysis

git checkout -b feature/improve-analysis

Edit analysis.py - add a new function at the bottom:

def calculate_seasonal_averages(df):
    """Group temperatures by season and calculate means."""
    seasons = {12: 'Summer', 1: 'Summer', 2: 'Summer',
               3: 'Autumn', 4: 'Autumn', 5: 'Autumn',
               6: 'Winter', 7: 'Winter', 8: 'Winter',
               9: 'Spring', 10: 'Spring', 11: 'Spring'}
    df = df.copy()
    df['season'] = df['date'].dt.month.map(seasons)
    return df.groupby(['station', 'season'])['temperature'].mean()

git add analysis.py
git commit -m "Add seasonal averages calculation"
git push origin feature/improve-analysis

Go to GitHub → open a Pull Request.

Person B: Update the documentation

git checkout -b feature/update-docs

Edit README.md - add a Methods section after "Project Overview":

## Methods

We use linear regression to detect long-term temperature trends and
seasonal decomposition to identify cyclical patterns. Data cleaning
removes stations with more than 20% missing values.

git add README.md
git commit -m "Add methods section to README"
git push origin feature/update-docs

Go to GitHub → open a Pull Request.

Both: Review and Merge

1. Go to your partner's Pull Request
2. Click Files changed - read what they did
3. Leave at least one comment on a specific line (click the + button next to a line)
4. Click Review changes → Approve → Submit review
5. Click Merge pull request

Both PRs should now be merged into main.

---

Round 2: The Merge Conflict (10 minutes)

Now you'll both edit the same line of the same file. This will create a merge conflict - and that's the point!

Both people:

git checkout main
git pull origin main

You should both now have the merged changes from Round 1.

Person A:

git checkout -b experiment/method-a

Edit line 1 of analysis.py to:

"""Analysis module - Method A: Linear Regression"""

git add analysis.py
git commit -m "Set analysis method to Linear Regression"
git push origin experiment/method-a

Go to GitHub, open a PR, and merge it immediately.

Person B:

git checkout -b experiment/method-b

Edit line 1 of analysis.py to:

"""Analysis module - Method B: Random Forest"""

git add analysis.py
git commit -m "Set analysis method to Random Forest"
git push origin experiment/method-b

Go to GitHub → open a PR.

You'll see a merge conflict! GitHub will say this branch has conflicts that must be resolved.

Person B: Resolve the conflict

# Pull the latest main into your branch
git checkout experiment/method-b
git pull origin main

Git will tell you there's a conflict. Open analysis.py - you'll see something like:

<<<<<<< HEAD
"""Analysis module - Method B: Random Forest"""
=======
"""Analysis module - Method A: Linear Regression"""
>>>>>>> main

Edit the file to keep what you want (or combine them!). Remove the <<<<<<<, =======, and >>>>>>> markers. For example:

"""Analysis module - Method B: Random Forest (Alternative to Linear Regression)"""

Then:

git add analysis.py
git commit -m "Resolve merge conflict: keep Random Forest method"
git push origin experiment/method-b

Go back to your PR on GitHub - it should now be mergeable!

---

Debrief Discussion

• How did the merge conflict feel?
• How would you prevent conflicts in a real project?
• What role do pull requests play in research collaboration?

Key takeaway: Merge conflicts aren't errors - they're Git asking you to make a decision. Communication with your collaborators prevents most of them.