Netflix Content Analysis

Exploratory data analysis of Netflix's content catalogue, including data cleaning, genre/country breakdowns, and a Power BI dashboard.

---

Project Structure

File	Description
raw_netflix.csv	Original dataset (8,807 rows, 12 columns) — unprocessed Netflix titles
cleaned_netflix.csv	Cleaned dataset (8,797 rows, 13 columns) — ready for analysis
netflix_country.csv	Exploded country table — one row per show-country pair
netflix_genres.csv	Exploded genre table — one row per show-genre pair
analysis.ipynb	Jupyter notebook covering cleaning steps and visualisations
Netflix_Dashboard.pbix	Power BI dashboard for interactive exploration

---

Dataset

Source: Netflix Titles dataset (publicly available on Kaggle)

Raw columns: show_id, type, title, director, cast, country, date_added, release_year, rating, duration, listed_in, description

Added during cleaning: year_added (extracted from date_added)

---

Data Cleaning (analysis.ipynb)

The notebook performs the following steps on raw_netflix.csv:

• Missing values — filled director, cast, and country with placeholder values; imputed missing duration by content type ('90 min' for Movies, '1 Season' for TV Shows); filled missing rating as 'NR'
• Dropped rows — removed records where date_added was null (10 rows removed)
• Date parsing — stripped whitespace and converted date_added to datetime; extracted year_added
• Rating fixes — corrected misplaced duration values ('74 min', '84 min', '66 min') in the rating column to 'NR'; standardised 'Not Rated' and 'UR' → 'NR'
• Text cleanup — stripped leading/trailing whitespace from all text columns; removed embedded newlines from title, description, cast, director, country, listed_in, duration, rating
• Deduplication — dropped exact duplicate rows
• Column names — standardised to lowercase with underscores

---

Derived Tables

Because country and listed_in (genres) can contain multiple comma-separated values per title, two normalised tables were created by exploding these columns:

• netflix_country.csv — show_id + one country per row
• netflix_genres.csv — show_id + one genre per row

These are used in the Power BI dashboard for accurate country and genre counts.

---

Visualisations

The notebook generates matplotlib charts styled with a Netflix dark theme (#141414 background, #E50914 red palette). Charts cover content type splits, top genres, content additions over time, and rating distributions.

The Netflix_Dashboard.pbix file provides an interactive Power BI dashboard built on the cleaned and derived tables.

To open the dashboard: Power BI Desktop is required (free download from Microsoft).

---

Requirements

pandas
matplotlib
jupyter

Install with:

pip install pandas matplotlib jupyter

---

Usage

1. Open analysis.ipynb in Jupyter and run all cells to reproduce the cleaning steps and charts.
2. The cleaned outputs (cleaned_netflix.csv, netflix_country.csv, netflix_genres.csv) are already included.
3. Open Netflix_Dashboard.pbix in Power BI Desktop for the interactive dashboard.

Dashboard Preview

Charts