What is Data Science?

What a data scientist actually does

A data scientist takes raw, messy data and turns it into an answer someone can act on. The boss does not want a spreadsheet — they want to know “which product should we stock more of?” or “why did sales drop in March?”. Your job is to find that answer in the data.

It is detective work. The data is the pile of clues. You clean it up, look for patterns, draw a chart that makes the pattern obvious, and write down what you found.

The data-science workflow

Almost every project follows the same five steps, in order. You will repeat this loop again and again.

1. Ask a clear question (“Do bigger houses really sell for more?”).
2. Get the data (a CSV file, a database, a website).
3. Clean it — fix missing values, wrong types and duplicates.
4. Explore & visualise — summarise the numbers and draw charts.
5. Conclude — say what you found, in plain language.

A tiny taste

Here is the whole workflow in spirit — load a little data, ask one question, and answer it.

import pandas as pd

# Tiny dataset: city and average house price (in thousands)
data = {'city': ['Pune', 'Pune', 'Delhi', 'Delhi'],
        'price': [60, 80, 120, 140]}
df = pd.DataFrame(data)

# Question: what is the average price per city?
print(df.groupby('city')['price'].mean())