Understanding describe() in Pandas: A Beginner-Friendly Guide

When working with data in Python using Pandas, one of the first things you usually want to know is:

❓ What does my data look like? What are the averages, minimums, and maximums?

Pandas provides a powerful method called describe() that helps you answer these questions instantly.

In this tutorial, you’ll learn:

• What describe() does
• Why pd.describe() does not work
• How to correctly use describe()
• Common beginner mistakes
• Practical examples

---

✅ 1. Setting Up the Context

Imagine you are working with a dataset that contains real estate information, such as:

• Property area in square meters (area_m2)
• Property price in US dollars (price_usd)

Let’s start by importing Pandas and creating a sample DataFrame.

Initial Code

import pandas as pd

# Create sample data
data = {
    "area_m2": [60, 75, 90, 120, 150],
    "price_usd": [70000, 95000, 120000, 170000, 230000]
}

# Create DataFrame
df = pd.DataFrame(data)

# View data
print(df)

Output

   area_m2  price_usd
0       60      70000
1       75      95000
2       90     120000
3      120     170000
4      150     230000

Now we have real data to work with.

---

❌ 2. A Common Beginner Mistake

Many beginners try something like this:

stats = pd.describe(["area_m2"], ["price_usd"])

This does not work.

Why?

Because:

1. describe() is not a function inside pd
2. You passed column names as strings, not actual data
3. describe() works on DataFrames or Series

So Pandas cannot calculate statistics from this.

---

✅ 3. The Correct Way to Use describe()

In Pandas, describe() is a method of a DataFrame.

That means:

You must call it on df, not on pd.

Correct Code

stats = df[["area_m2", "price_usd"]].describe()
print(stats)

---

📊 4. Understanding the Output

You will get something like:

          area_m2     price_usd
count     5.0000        5.0000
mean     99.0000   137000.0000
std      35.4965    63529.9600
min      60.0000    70000.0000
25%      75.0000    95000.0000
50%      90.0000   120000.0000
75%     120.0000   170000.0000
max     150.0000   230000.0000

Let’s understand what each row means:

Statistic	Meaning
count	Number of values
mean	Average
std	Standard deviation
min	Smallest value
25%	First quartile
50%	Median
75%	Third quartile
max	Largest value

These give you a quick summary of your dataset.

---

✅ 5. Describing a Single Column

If you want statistics for only one column:

For area_m2

area_stats = df["area_m2"].describe()
print(area_stats)

For price_usd

price_stats = df["price_usd"].describe()
print(price_stats)

This returns statistics for just that column.

---

✅ 6. Describing All Numeric Columns

If your dataset has many columns, you can simply write:

stats = df.describe()
print(stats)

Pandas will automatically:

• Select numeric columns
• Ignore text columns
• Summarize useful statistics

This is very helpful in real-world projects.

---

⚠️ 7. Why Passing Column Names Does Not Work

Look at this incorrect example:

df.describe(["area_m2", "price_usd"])

This does not work because:

• "area_m2" is just text
• Pandas needs actual values
• Not column labels

Think like this:

❌ Column names = labels
✅ Column data = numbers

describe() works only with numbers.

---

💡 8. The Pattern You Should Remember

Always remember this structure:

For multiple columns:

df[["col1", "col2"]].describe()

For one column:

df["col1"].describe()

For whole DataFrame:

df.describe()

If you remember this pattern, you’ll avoid most mistakes.

---

🌟 9. Real-World Use Case

In real projects, describe() is often used:

• Before cleaning data
• Before building ML models
• While exploring datasets
• To detect outliers
• To check data quality

Example:

# Check dataset before analysis
print(df.describe())

This helps you understand your data in seconds.

---

📌 10. Summary

Let’s summarize what we learned.

❌ Wrong Way

pd.describe(["area_m2"], ["price_usd"])

✅ Correct Ways

# Multiple columns
df[["area_m2", "price_usd"]].describe()

# Single column
df["area_m2"].describe()

# Entire DataFrame
df.describe()

---

✅ Key Takeaways

✔ describe() is a DataFrame method
✔ It summarizes numeric data
✔ It does not work on column names
✔ It helps in quick data exploration

---