IMDB Analysis: Serbian and Yugoslav Films

Analysis of movies from Serbia and Yugoslavia across history using Python (Pandas & Matplotlib). The goal is to explore trends, top-rated movies and directors, and the most popular genres by decade.

Data

This project uses the IMDB Extensive Dataset from Kaggle.
Download the dataset and place the movie-level CSV in the project root as IMDB.csv (or adjust the script path).

Technologies

• Python 3
• Pandas (data processing)
• Matplotlib (visualization)

Main Analysis Steps

1. Load & clean data
   • Remove unused columns (budget, description, etc.)
   • Filter movies from Serbia/Yugoslavia
   • Convert numeric fields (year, ratings, votes)
2. Director analysis
   • Film count per director
   • Average rating per director (for those with more than 3 films)
   • Top 10 directors by average rating
3. Decade analysis
   • Assign a decade to each film
   • Split multi-valued genres (films often have several)
   • Build a pivot table: number of films per genre × decade
4. Visualization
   • Horizontal bar chart: Top 10 directors by rating

Results

The analysis produces the following outputs:

• Genre × Decade table → saved as pivot_genre_by_decade.csv.

  • Rows are decades (e.g., 1960, 1970, …), columns are genres (Drama, Comedy, …).
  • Each cell is the count of films tagged with that genre in that decade.
  • Note: a film with multiple genres contributes once to each of its listed genres.

• Directors summary (printed to console) → broj_po_reziseru3 table:

  • director — director name
  • broj_filmova — number of films
  • prosek_ocena — average IMDB rating
  • broj_glasova — total votes
  • najraniji_film / najkasniji_film — earliest/latest film year
  • period_stvaranja — active span (latest - earliest)
  • Filtered to directors with > 3 films, sorted by prosek_ocena.

• Top 10 Directors chart → displayed in a window:

  • Horizontal bar chart of the Top 10 directors by average rating (after the “> 3 films” filter).
  • You can optionally save it (e.g., outputs/top10_directors.png).

• Quick console sanity checks:

  • Year range of the dataset (min/max).
  • Top-rated titles (largest avg_vote).
  • Films with avg_vote > 8.1 (sorted by rating, then votes).

Interpretation: Use pivot_genre_by_decade.csv to see which genres dominate each decade.
Use the directors summary and chart to spot consistently high-rated directors with a meaningful body of work.

Outputs

• pivot_genre_by_decade.csv — saved automatically.
• “Top 10 directors” chart — displayed (optionally saved as outputs/top10_directors.png).

How to Run

1. Place IMDB.csv next to your Python script (or notebook).
2. Install dependencies:

   pip install pandas matplotlib

3. Run the analysis (script version):

   python imdb_analysis.py

   A CSV with the genre × decade pivot will be saved, and a chart of the Top 10 directors will be shown.

Notes

• Uses a local IMDB.csv from the Kaggle dataset (not included).
• Deduplicates by imdb_title_id.
• country filter allows mixed entries (e.g., “Serbia, France”).
• Genres are split by , ; multi-genre films count once per genre.
• Outputs: saves pivot_genre_by_decade.csv and shows a Top-10 directors chart.

Suggested Files

IMDB.csv
imdb_analysis.py
outputs/            # optional (for saved CSV/PNG)
README.md

License

MIT (or your preferred license).