🔎Exploratory Data Analysis · Project
Combine Pandas + charts to interrogate a real dataset.
📖 Key concepts
- Ask questions of the data
- Clean: missing values, types
- Summarise: describe(), value_counts()
- Visualise to confirm
💻 Try this code
import pandas as pd
df = pd.read_csv("data.csv")
print(df.shape)
print(df.describe())
print(df["category"].value_counts())
# then plot the interesting columns👥 Learn it your way
🧒 Young Explorers
Be a data detective: ask a question, then let the data answer it.
Pick a fun dataset (e.g., Pokémon, movies) and answer 3 questions about it.
💼 Professionals
Build a repeatable EDA checklist: shape, dtypes, nulls, distributions, correlations.
On a real CSV, report 3 insights with supporting summaries/plots.
EDA is how you decide which features matter and whether the data is fair/representative — the groundwork that determines if a model will succeed or fail.
Choose a real CSV dataset. Load it with Pandas, clean it, answer 3–5 questions with summaries and at least 2 charts, and write a short "what I found" note. Pro stretch: identify a potential bias or data-quality issue.
✅ Quick check
What does value_counts() give you?
How many times each unique value appears in a column.
Why do EDA before modelling?
To understand, clean, and trust the data first.