added table of stats

README.md

@@ -15,6 +15,30 @@ The python package `pyodbc` was then used to load the tables from this database
 ### Results
 
 ![Rating Distribution](./figures/rating_distribution.png)
+
+| genre     |   median |   top_20perc |   num_films |
+|:----------|---------:|-------------:|------------:|
+| Horror    |  6.125   |         6.8  |        1103 |
+| Sci-Fi    |  6.356   |         7.2  |         644 |
+| Family    |  6.37059 |         7.3  |         420 |
+| Fantasy   |  6.4     |         7.3  |         693 |
+| Action    |  6.41585 |         7.2  |        2041 |
+| Thriller  |  6.48305 |         7.3  |        1458 |
+| Mystery   |  6.50278 |         7.4  |         913 |
+| Comedy    |  6.52558 |         7.3  |        3244 |
+| Adventure |  6.61695 |         7.4  |        1500 |
+| Romance   |  6.7     |         7.4  |        1483 |
+| Sport     |  6.73333 |         7.4  |         171 |
+| Crime     |  6.73429 |         7.4  |        1741 |
+| Animation |  6.96471 |         7.7  |         426 |
+| Drama     |  7.00262 |         7.6  |        4790 |
+| Musical   |  7.15    |         7.6  |          87 |
+| Biography |  7.175   |         7.62 |         585 |
+| History   |  7.2     |         7.7  |         295 |
+| Western   |  7.3     |         7.7  |         105 |
+| War       |  7.4     |         8    |         200 |
+
+
 ---
 
 To do:

SQLQuery.sql

@@ -152,10 +152,68 @@ FROM IMDBRatings..TitleRatings AS tr
 		ON tb.tconst = te.imdb_title_id
 WHERE (titleType = 'movie' OR titleType = 'tvMovie')
 AND te.country NOT LIKE '%India%'
-AND numVotes > 10000
+AND numVotes > 10000;
 -- ORDER BY averageRating ASC, numVotes ASC;
 
 
+-- Exploring AVG values between media types
+SELECT DISTINCT tb.titleType
+FROM IMDBRatings..TitleRatings AS tr
+	LEFT JOIN IMDBRatings..TitleBasics AS tb
+		ON tr.tconst = tb.tconst
+	LEFT JOIN IMDBRatings..TitleExtras AS te
+		ON tb.tconst = te.imdb_title_id
+WHERE numVotes > 10000;
+--Result: 
+	--video
+	--tvShort
+	--tvMovie
+	--videoGame
+	--short
+	--movie
+	--tvSeries
+	--tvMiniSeries
+	--tvSpecial
+	--tvEpisode
+
+SELECT AVG(tr.averageRating)
+FROM IMDBRatings..TitleRatings AS tr
+	LEFT JOIN IMDBRatings..TitleBasics AS tb
+		ON tr.tconst = tb.tconst
+	LEFT JOIN IMDBRatings..TitleExtras AS te
+		ON tb.tconst = te.imdb_title_id
+WHERE numVotes > 10000
+AND (tb.titleType = 'movie' OR tb.titleType = 'tvMovie');
+-- Result: 6.65
+
+SELECT AVG(tr.averageRating)
+FROM IMDBRatings..TitleRatings AS tr
+	LEFT JOIN IMDBRatings..TitleBasics AS tb
+		ON tr.tconst = tb.tconst
+	LEFT JOIN IMDBRatings..TitleExtras AS te
+		ON tb.tconst = te.imdb_title_id
+WHERE numVotes > 10000
+AND (tb.titleType = 'tvSeries' OR tb.titleType = 'tvMiniSeries');
+-- Result: 7.73
+
+-- Normally, episodes receive ~10x less votes than the show rating
+-- This approximately gathers episodes of comparable popularity
+SELECT AVG(tr.averageRating)
+FROM IMDBRatings..TitleRatings AS tr
+	LEFT JOIN IMDBRatings..TitleBasics AS tb
+		ON tr.tconst = tb.tconst
+	LEFT JOIN IMDBRatings..TitleExtras AS te
+		ON tb.tconst = te.imdb_title_id
+WHERE numVotes > 1000
+AND tb.titleType = 'tvEpisode';
+-- Result: 8.15
+
+
+
+
+
+
+
 
 
 

rating_plots.py

@@ -34,7 +34,6 @@
 cursor = cnxn.cursor()
 
 # Creating genres list
-
 genres = [
     'Action',
     'Adventure',
@@ -66,42 +65,50 @@
 db = pd.read_sql(ratings_sql, cnxn)
 
 # Creating Table of Stats
-
-medianVals = {'genre': [], 
-              'median': []}
+genreVals = {'genre': [], 
+              'median': [],
+              'top_20perc': [],
+              'num_films': []}
 for genre in genres:
     genre_db = db[db.genres.str.contains(genre, regex=True, na=False)]
+    num_films = len(genre_db)
+    top_20perc = genre_db.averageRating.quantile(q=0.8)
     genre_db['movierank'] = ((np.arange(len(genre_db))+1)/len(genre_db)*100).round(2)
     median = genre_db[genre_db.movierank<52][genre_db.movierank>48].averageRating.mean()
-    medianVals['genre'].append(genre)
-    medianVals['median'].append(median)
-
-medianVals = pd.DataFrame(medianVals).sort_values('median')
-
+    genreVals['genre'].append(genre)
+    genreVals['median'].append(median)
+    genreVals['top_20perc'].append(top_20perc)
+    genreVals['num_films'].append(num_films)
+
+genreVals = pd.DataFrame(genreVals).sort_values('median')
+genreVals_table = tabulate(genreVals, 
+                           headers='keys', 
+                           tablefmt='pipe', 
+                           showindex='never')
+
+# Creating plot with custom 
 NUM_COLORS = len(genres)
-
 cm = plt.get_cmap('gist_rainbow')
 fig = plt.figure()
 ax = fig.add_subplot(111)
 ax.set_prop_cycle('color', [cm(1.*i/NUM_COLORS) for i in range(NUM_COLORS)])
-
-for genre in medianVals.genre:
+for genre in genreVals.genre:
     genre_db = db[db.genres.str.contains(genre, regex=True, na=False)]
     genre_db['movierank'] = ((np.arange(len(genre_db))+1)/len(genre_db)*100).round(2)
     genre_db = genre_db.groupby('averageRating')['movierank'].mean()
     genre_db = genre_db.reset_index()
     ax.plot(genre_db.averageRating, genre_db.movierank, label=genre, alpha=0.4)
 
-g1 = ax.grid(b=True, which='major', linestyle='-', linewidth=0.5)
-g2 = ax.grid(b=True, which='minor', linestyle='-', linewidth=0.2)
+g1 = ax.grid(b=True, which='major', linestyle='-')
+g2 = ax.grid(b=True, which='minor', linestyle='-', linewidth=0.4)
 ax.minorticks_on()
 plt.legend(loc='center left', bbox_to_anchor=(1.05, 0.5))
 plt.ylabel('Percent Rank')
 plt.xlabel('User Rating')
 plt.xticks(np.arange(0, 11, 1))
 plt.yticks(np.arange(0, 101, 10))
 plt.title('Rating Distribution')
-plt.savefig('rating_distribution.png', dpi=600, bbox_inches='tight')
+plt.savefig('rating_distribution.png', dpi=200, bbox_inches='tight')