Module 02: Pandas Introduction - Participation Exercises

Open In Colab

Module 02: Pandas Introduction - Participation Exercises#

Exercise 2.1: Prediction - Data Types#

Type: 🔮 Prediction (3 min)

You receive a CSV file with experimental data. Before loading it, predict what data types pandas will assign:

# Imagine this CSV content:
csv_content = """
experiment_id,temperature,pressure,catalyst,yield,date,notes
EXP001,350.5,1.2,Pt/Al2O3,78.3,2024-01-15,Good run
EXP002,375.0,1.5,Pd/C,82.1,2024-01-16,
EXP003,400.0,1.8,Pt/Al2O3,NaN,2024-01-17,Equipment failure
"""

# Predict the dtype for each column:
# experiment_id: ???
# temperature: ???
# pressure: ???
# catalyst: ???
# yield: ???
# date: ???
# notes: ???

Your dtype predictions:

Column

Predicted dtype

Reasoning

experiment_id

temperature

pressure

catalyst

yield

date

notes

Exercise 2.2: Mini-Exercise - Data Exploration Race#

Type: 🔧 Mini-Exercise (7 min)

Load the dataset below and answer all questions as quickly as possible. First person done raises their hand!

import pandas as pd
import numpy as np

# Create sample reactor data
np.random.seed(42)
n = 100
df = pd.DataFrame({
    'reactor': np.random.choice(['R1', 'R2', 'R3'], n),
    'temperature': np.random.uniform(300, 500, n),
    'pressure': np.random.uniform(1, 10, n),
    'conversion': np.random.uniform(0.3, 0.95, n),
    'shift': np.random.choice(['Day', 'Night'], n)
})
df.loc[np.random.choice(n, 5), 'conversion'] = np.nan  # Add some missing values

# QUESTIONS - Answer using pandas operations:
# 1. How many rows and columns?
# 2. How many missing values in 'conversion'?
# 3. What is the mean temperature for reactor R2?
# 4. How many experiments were run on the Night shift?
# 5. What is the maximum conversion?

# Your code here:

Exercise 2.3: Discussion - Missing Data Strategies#

Type: 💬 Discussion (5 min)

You have sensor data from a reactor with 10% missing values. Discuss with a partner:

  1. What are three different ways to handle the missing data?

  2. What are the pros and cons of each approach?

  3. Does it matter why the data is missing?

Scenario: The missing values occur mostly during shift changes. How does this affect your strategy?

Discussion notes:

Strategy

Pros

Cons

Contents