How to fix FileNotFoundError in Jupyter Notebook & Google Colab

You’re sure the CSV or image is in the folder. But when you run the cell, Jupyter or Colab says:

FileNotFoundError: [Errno 2] No such file or directory: 'data/train.csv'

This is one of the most confusing early errors in data science – mostly because it’s not about Python at all, it’s about **where your notebook is looking from**.

Step 1: Understand “working directory” vs “file location”

Two things matter:

• File location: where your file actually exists on disk.
• Working directory: the folder Python is treating as “current” when you run open() or pd.read_csv().

If those two don’t match, you get FileNotFoundError – even if you can see the file in your file explorer.

Step 2: Fix FileNotFoundError in local Jupyter Notebook

1. Print your current working directory

In a new cell, run:

import os
print(os.getcwd())

This shows the folder Jupyter is currently running in. Compare it to where your file actually lives.

2. Use project-relative paths, not absolute desktop paths

Instead of this:

df = pd.read_csv("C:/Users/me/Desktop/data/train.csv")

Prefer a project structure like:

project/
  notebook.ipynb
  data/
    train.csv
    test.csv

And use:

df = pd.read_csv("./data/train.csv")

3. Move the notebook or change the working directory

If your notebook is inside a notebooks/ folder, but your data is up one level, you have two options:

• Move the notebook to the project root, or
• Change working directory at the top of the notebook:

import os
os.chdir("..")  # go up one folder

Step 3: Fix FileNotFoundError in Google Colab

1. Confirm where your files live in Drive

Typical flow:

1. Mount Drive:

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

2. Inspect the path in the left-hand file browser (e.g. /content/drive/MyDrive/colab-data/train.csv).
3. Use that path (or a relative one) in your code.

2. Avoid hard-coded “Downloads” or random paths

Dragging a file into Colab’s temporary /content file space works until the runtime resets. Prefer a stable folder under Drive, like:

/content/drive/MyDrive/projects/my-ml-project/
  notebook.ipynb
  data/
    train.csv

3. Use os.listdir() to debug

If you’re not sure what’s in a folder, list it from the notebook:

import os
print(os.listdir())
print(os.listdir("./data"))

If train.csv doesn’t appear there, Python is correct: the file isn’t where the path says it is.

Step 4: Make your data paths stable and shareable

Fixing the error for yourself is step one. The next step is making sure your **future self and collaborators** don’t hit the same FileNotFoundError.

1. Standardise your project layout

Adopt a simple pattern for all your ML projects:

project-name/
  notebook.ipynb
  data/
    raw/
    processed/
  models/
  capsules/           # NoteCapsule snapshots

2. Use NoteCapsule to capture data layout

NoteCapsule can create a Capsule that includes a data_manifest.json file describing which data paths your notebook expects.

!pip install notebookcapsule -q

from notebookcapsule import create_capsule

create_capsule(
    name="after-fixing-filenotfounderror",
    data_dirs=["./data"]
)

That way, anyone opening your project later can see exactly which files need to exist and where.

Quick checklist for fixing FileNotFoundError