Students Performance Profiling
Load the Students Performance dataset, profile it thoroughly, and formulate EDA questions for the final report.
🎯 What you'll learn:
- Load the bundled Students Performance dataset and verify data integrity
- Profile the dataset using the complete workflow from Module 1
- Formulate primary, secondary, and tertiary EDA questions for the capstone report
- Identify data quality issues and plan the analysis strategy
This is the start of your capstone project. You will load the Students Performance in Exams dataset, profile it using the systematic workflow from Module 1, and formulate the EDA questions that will guide your final report. The work you do here determines the quality of everything that follows.
Key Concepts
Loading the dataset
The Students Performance dataset contains 1000 students with 8 features: gender, race/ethnicity, parental education, lunch type, test preparation course, and scores in math, reading, and writing.
python
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
# students-performance.csv ships with the course — load it from the browser file system.
df = pd.read_csv("students-performance.csv")
print(f"Dataset loaded: {df.shape[0]} rows, {df.shape[1]} columns")
print(f"\nColumns: {list(df.columns)}")
print(f"\nFirst 5 rows:")
df.head()Complete profiling workflow
Run the full profiling report using the function from Module 1:
python
def profile_dataset(df, name="Dataset"):
"""Generate a complete profiling report."""
print(f"{'='*60}")
print(f" PROFILING REPORT: {name}")
print(f"{'='*60}")
print(f"\nSTRUCTURE")
print(f" Rows: {df.shape[0]:,}")
print(f" Columns: {df.shape[1]}")
print(f" Memory: {df.memory_usage(deep=True).sum() / 1e6:.2f} MB")
print(f"\nCOLUMN TYPES")
print(df.dtypes.value_counts().to_string())
missing = df.isnull().sum()
if missing.any():
print(f"\nMISSING VALUES")
for col in missing[missing > 0].index:
pct = missing[col] / len(df) * 100
print(f" {col}: {missing[col]:,} ({pct:.1f}%)")
else:
print(f"\nMISSING VALUES: None detected")
n_dupes = df.duplicated().sum()
print(f"\nDUPLICATES: {n_dupes} rows ({n_dupes/len(df)*100:.1f}%)")
num_df = df.select_dtypes(include="number")
if not num_df.empty:
print(f"\nNUMERICAL SUMMARY")
print(num_df.describe().round(2).to_string())
cat_df = df.select_dtypes(include="object")
if not cat_df.empty:
print(f"\nCATEGORICAL SUMMARY")
for col in cat_df.columns:
n_unique = df[col].nunique()
print(f" {col}: {n_unique} unique — {df[col].value_counts().to_dict()}")
print(f"\n{'='*60}")
profile_dataset(df, "Students Performance in Exams")Detailed categorical analysis
python
cat_cols = ["gender", "race/ethnicity", "parental level of education",
"lunch", "test preparation course"]
for col in cat_cols:
print(f"\n{'─'*40}")
print(f" {col.upper()}")
print(f"{'─'*40}")
vc = df[col].value_counts()
pct = df[col].value_counts(normalize=True).round(3)
summary = pd.DataFrame({"count": vc, "proportion": pct})
print(summary)Detailed numerical analysis
python
num_cols = ["math score", "reading score", "writing score"]
for col in num_cols:
print(f"\n{'─'*40}")
print(f" {col.upper()}")
print(f"{'─'*40}")
s = df[col]
print(f" Mean: {s.mean():.2f}")
print(f" Median: {s.median():.2f}")
print(f" Std: {s.std():.2f}")
print(f" Min: {s.min()}")
print(f" Max: {s.max()}")
print(f" Skew: {s.skew():.3f}")
print(f" Kurt: {s.kurtosis():.3f}")
print(f" IQR: {s.quantile(0.75) - s.quantile(0.25):.2f}")
print(f" Outliers (IQR): {((s < s.quantile(0.25) - 1.5 * (s.quantile(0.75) - s.quantile(0.25))) | (s > s.quantile(0.75) + 1.5 * (s.quantile(0.75) - s.quantile(0.25)))).sum()}")Visual profiling summary
python
fig, axes = plt.subplots(2, 3, figsize=(15, 10))
# Numerical distributions
for i, col in enumerate(num_cols):
sns.histplot(df[col], kde=True, ax=axes[0, i], bins=20, color="steelblue")
mean_val = df[col].mean()
axes[0, i].axvline(mean_val, color="red", linestyle="--", label=f"Mean: {mean_val:.1f}")
axes[0, i].set_title(col.replace(" score", " Scores").title())
axes[0, i].legend(fontsize=9)
# Categorical counts
cat_to_plot = ["gender", "race/ethnicity", "lunch"]
for i, col in enumerate(cat_to_plot):
sns.countplot(data=df, x=col, ax=axes[1, i], palette="Set2")
axes[1, i].set_title(col.title())
axes[1, i].tick_params(axis="x", rotation=45)
plt.suptitle("Students Performance — Visual Profile", fontsize=14, fontweight="bold")
plt.tight_layout()
plt.show()Formulating EDA questions
Based on the profiling, formulate your analytical questions:
python
eda_questions = {
"primary": "What factors most strongly predict student performance on exams?",
"secondary": [
"How do math, reading, and writing scores distribute across the population?",
"Do gender and ethnicity interact to produce score differences?",
"What is the magnitude of the test preparation effect?",
"Does lunch type (proxy for socioeconomic status) predict scores?",
"How does parental education level relate to student performance?",
],
"tertiary": [
"Are the three scores correlated? Is the correlation the same across genders?",
"Do outliers represent data entry errors or genuine high/low performers?",
"Is there an interaction between test preparation and lunch type?",
"Which subgroup shows the largest performance gap?",
],
"analysis_plan": [
"Step 1: Univariate analysis of all 8 variables",
"Step 2: Bivariate analysis — scores vs each categorical variable",
"Step 3: Correlation analysis of the three score variables",
"Step 4: Multivariate analysis — interactions between factors",
"Step 5: Advanced visualizations and storytelling",
"Step 6: Final report with findings and recommendations",
]
}
print("PRIMARY QUESTION:")
print(f" {eda_questions['primary']}\n")
print("SECONDARY QUESTIONS:")
for i, q in enumerate(eda_questions["secondary"], 1):
print(f" {i}. {q}")
print("\nTERTIARY QUESTIONS:")
for i, q in enumerate(eda_questions["tertiary"], 1):
print(f" {i}. {q}")
print("\nANALYSIS PLAN:")
for step in eda_questions["analysis_plan"]:
print(f" {step}")Try It
Run the full profiling workflow on the Students Performance dataset. Answer these questions from your profiling output:
python
# Complete the profiling and answer:
questions = {
"q1": "How many students are in the dataset?",
"q2": "What are the column names?",
"q3": "Are there any missing values?",
"q4": "How many unique values does each categorical column have?",
"q5": "What is the mean math score?",
"q6": "What is the standard deviation of reading scores?",
"q7": "Which categorical column has the most balanced distribution?",
"q8": "Are there any numerical outliers in the score columns?",
}
# Your answers:
answers = {
"q1": "1000 students",
"q2": "gender, race/ethnicity, parental level of education, lunch, test preparation course, math score, reading score, writing score",
"q3": "No missing values",
"q4": "gender: 2, race/ethnicity: 5, parental level of education: 6, lunch: 2, test preparation course: 2",
"q5": "66.1",
"q6": "14.6",
"q7": "gender (50/50 split) or lunch (roughly 65/35 split)",
"q8": "Yes — some students scored below 30 on math (potential outliers)",
}
for q, a in answers.items():
print(f"{questions[q]}\n → {a}\n")Key Takeaways
- Always load data from a reproducible source (URL) so others can replicate your analysis
- The profiling workflow from Module 1 is your standard starting point for any dataset
- Formulate questions before analysis, they keep you focused and prevent scope creep
- The Students Performance dataset is clean (no missing values) but has interesting categorical variables to explore
- Your analysis plan should flow logically: univariate → bivariate → multivariate → storytelling
Practice Challenge
Write a Python script that loads the dataset, profiles it, and generates a summary dictionary with: row_count, column_count, missing_columns, score_means, score_stds, categorical_cardinalities, and eda_questions. Print the dictionary formatted.
🧩 Challenge, think first, then reveal
python
import pandas as pd
# students-performance.csv ships with the course — load it from the browser file system.
df = pd.read_csv("students-performance.csv")
score_cols = ["math score", "reading score", "writing score"]
cat_cols = ["gender", "race/ethnicity", "parental level of education",
"lunch", "test preparation course"]
summary = {
"row_count": len(df),
"column_count": len(df.columns),
"missing_columns": {col: int(df[col].isnull().sum()) for col in df.columns if df[col].isnull().sum() > 0},
"score_means": {col: round(df[col].mean(), 2) for col in score_cols},
"score_stds": {col: round(df[col].std(), 2) for col in score_cols},
"categorical_cardinalities": {col: df[col].nunique() for col in cat_cols},
"eda_questions": {
"primary": "What factors most strongly predict student performance?",
"secondary": [
"How do scores distribute across subjects?",
"Do demographic factors correlate with performance gaps?",
"What is the effect of test preparation course completion?",
],
},
}
for key, value in summary.items():
print(f"\n{key.upper().replace('_', ' ')}:")
if isinstance(value, dict):
for k, v in value.items():
print(f" {k}: {v}")
elif isinstance(value, list):
for item in value:
print(f" - {item}")
else:
print(f" {value}") ✅ Quick check +10 XP
1. What should you do before starting any analysis on a new dataset?
2. How many rows does the Students Performance dataset have?
3. What is the first thing you should check after loading a dataset?