Google Colab Walkthrough

What is Google Colab?

Google Colab (Colaboratory) is a free, cloud-hosted Jupyter Notebook environment.

It allows you to:

Run Python in the browser (no local install)
Use free GPUs/TPUs (great for ML workloads)
Share notebooks easily
Connect to Google Drive

For data analytics, Colab is excellent when:

Your machine is slow
You want a quick setup
You want to share a notebook link with others

Create a new notebook

Go to: https://colab.research.google.com/
Sign in with a Google account
Click New notebook

A notebook opens with a default code cell.

Understanding Colab interface

Key menus you’ll use:

Runtime: run cells, restart runtime, change hardware
File: save a copy, download .ipynb, download .py
Edit: find/replace, shortcuts

Colab runs notebooks on remote servers called a runtime.

Run your first code

In a code cell:

hello

print("Hello from Colab")

Run with Shift + Enter.

Uploading datasets

You have a few common options.

Option 1: Upload directly (temporary)

Left sidebar → Files
Click Upload and choose your CSV

Important:

Uploaded files are stored in the runtime temporary disk.
If the runtime resets, uploaded files are gone.

Option 2: Google Drive (recommended)

Mount your Drive:

mount-drive

from google.colab import drive
drive.mount('/content/drive')

Then access files like:

read-csv

import pandas as pd
path = "/content/drive/MyDrive/data/titanic.csv"
df = pd.read_csv(path)
df.head()

Option 3: Download from a URL

download

import pandas as pd
url = "https://raw.githubusercontent.com/mwaskom/seaborn-data/master/titanic.csv"
df = pd.read_csv(url)
df.head()

Installing packages in Colab

Colab has many libraries installed, but not everything.

Use pip inside a notebook:

pip-install

!pip install yfinance

Or make it a bit quieter:

pip-install-q

!pip -q install yfinance

Changing runtime hardware (GPU/TPU)

Runtime → Change runtime type
Set:
- Hardware accelerator: GPU (or TPU)

You can confirm GPU availability:

check-gpu

import torch
print("CUDA available:", torch.cuda.is_available())

(If you don’t use PyTorch, you can also check with other tools.)

Saving and exporting your work

Save to Drive

File → Save a copy in Drive

Download notebook

File → Download → .ipynb

Export to Python

File → Download → .py

Visualize it

A Colab notebook is the same execute-explore experience as local Jupyter, just running on a remote machine.

diagram Colab runtime lifecycle mermaid

  flowchart LR
  A["Open colab.research.google.com"] --> B["New notebook"]
  B --> C["Connect to a runtime (CPU/GPU/TPU)"]
  C --> D["Mount Drive or upload data"]
  D --> E["Run cells with Shift+Enter"]
  E --> F["Save to Drive / download .ipynb"]

From opening a new notebook to a connected, ready-to-run cloud runtime

Colab tips for data analytics

Quickly preview data

preview

import pandas as pd
import numpy as np
 
df = pd.DataFrame({
    "city": ["Delhi", "Mumbai", "Pune"],
    "sales": [120, 90, 150]
})
 
df

Display full columns/rows (Pandas)

pandas-options

import pandas as pd
pd.set_option("display.max_columns", None)
pd.set_option("display.max_rows", 100)

Continue to: Virtual Environments for Data Science to understand why environments matter and how to maintain clean project dependencies.