1675 lines
60 KiB
Plaintext
1675 lines
60 KiB
Plaintext
{
|
||
"cells": [
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 57,
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import pandas as pd\n",
|
||
"import numpy as np"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 58,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>ONE-LINE</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>RunTime</th>\n",
|
||
" <th>Gross</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>\\nAction, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>\\nA woman with a mysterious illness is forced ...</td>\n",
|
||
" <td>\\n Director:\\nPeter Thorwarth\\n| \\n Star...</td>\n",
|
||
" <td>21,062</td>\n",
|
||
" <td>121.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>1</th>\n",
|
||
" <td>Masters of the Universe: Revelation</td>\n",
|
||
" <td>(2021– )</td>\n",
|
||
" <td>\\nAnimation, Action, Adventure</td>\n",
|
||
" <td>5.0</td>\n",
|
||
" <td>\\nThe war for Eternia begins again in what may...</td>\n",
|
||
" <td>\\n \\n Stars:\\nChris Wood, \\nSara...</td>\n",
|
||
" <td>17,870</td>\n",
|
||
" <td>25.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>2</th>\n",
|
||
" <td>The Walking Dead</td>\n",
|
||
" <td>(2010–2022)</td>\n",
|
||
" <td>\\nDrama, Horror, Thriller</td>\n",
|
||
" <td>8.2</td>\n",
|
||
" <td>\\nSheriff Deputy Rick Grimes wakes up from a c...</td>\n",
|
||
" <td>\\n \\n Stars:\\nAndrew Lincoln, \\n...</td>\n",
|
||
" <td>885,805</td>\n",
|
||
" <td>44.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>3</th>\n",
|
||
" <td>Rick and Morty</td>\n",
|
||
" <td>(2013– )</td>\n",
|
||
" <td>\\nAnimation, Adventure, Comedy</td>\n",
|
||
" <td>9.2</td>\n",
|
||
" <td>\\nAn animated series that follows the exploits...</td>\n",
|
||
" <td>\\n \\n Stars:\\nJustin Roiland, \\n...</td>\n",
|
||
" <td>414,849</td>\n",
|
||
" <td>23.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>4</th>\n",
|
||
" <td>Army of Thieves</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>\\nAction, Crime, Horror</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>\\nA prequel, set before the events of Army of ...</td>\n",
|
||
" <td>\\n Director:\\nMatthias Schweighöfer\\n| \\n ...</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR \\\n",
|
||
"0 Blood Red Sky (2021) \n",
|
||
"1 Masters of the Universe: Revelation (2021– ) \n",
|
||
"2 The Walking Dead (2010–2022) \n",
|
||
"3 Rick and Morty (2013– ) \n",
|
||
"4 Army of Thieves (2021) \n",
|
||
"\n",
|
||
" GENRE RATING \\\n",
|
||
"0 \\nAction, Horror, Thriller 6.1 \n",
|
||
"1 \\nAnimation, Action, Adventure 5.0 \n",
|
||
"2 \\nDrama, Horror, Thriller 8.2 \n",
|
||
"3 \\nAnimation, Adventure, Comedy 9.2 \n",
|
||
"4 \\nAction, Crime, Horror NaN \n",
|
||
"\n",
|
||
" ONE-LINE \\\n",
|
||
"0 \\nA woman with a mysterious illness is forced ... \n",
|
||
"1 \\nThe war for Eternia begins again in what may... \n",
|
||
"2 \\nSheriff Deputy Rick Grimes wakes up from a c... \n",
|
||
"3 \\nAn animated series that follows the exploits... \n",
|
||
"4 \\nA prequel, set before the events of Army of ... \n",
|
||
"\n",
|
||
" STARS VOTES RunTime Gross \n",
|
||
"0 \\n Director:\\nPeter Thorwarth\\n| \\n Star... 21,062 121.0 NaN \n",
|
||
"1 \\n \\n Stars:\\nChris Wood, \\nSara... 17,870 25.0 NaN \n",
|
||
"2 \\n \\n Stars:\\nAndrew Lincoln, \\n... 885,805 44.0 NaN \n",
|
||
"3 \\n \\n Stars:\\nJustin Roiland, \\n... 414,849 23.0 NaN \n",
|
||
"4 \\n Director:\\nMatthias Schweighöfer\\n| \\n ... NaN NaN NaN "
|
||
]
|
||
},
|
||
"execution_count": 58,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"df = pd.read_csv('movies.csv',header=0)\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 59,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>ONE-LINE</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>RunTime</th>\n",
|
||
" <th>Gross</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>A woman with a mysterious illness is forced in...</td>\n",
|
||
" <td>Director:Peter Thorwarth| Stars:Peri Baume...</td>\n",
|
||
" <td>21,062</td>\n",
|
||
" <td>121.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>1</th>\n",
|
||
" <td>Masters of the Universe: Revelation</td>\n",
|
||
" <td>(2021– )</td>\n",
|
||
" <td>Animation, Action, Adventure</td>\n",
|
||
" <td>5.0</td>\n",
|
||
" <td>The war for Eternia begins again in what may b...</td>\n",
|
||
" <td>Stars:Chris Wood, Sarah Michelle Gellar, Lena ...</td>\n",
|
||
" <td>17,870</td>\n",
|
||
" <td>25.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>2</th>\n",
|
||
" <td>The Walking Dead</td>\n",
|
||
" <td>(2010–2022)</td>\n",
|
||
" <td>Drama, Horror, Thriller</td>\n",
|
||
" <td>8.2</td>\n",
|
||
" <td>Sheriff Deputy Rick Grimes wakes up from a com...</td>\n",
|
||
" <td>Stars:Andrew Lincoln, Norman Reedus, Melissa M...</td>\n",
|
||
" <td>885,805</td>\n",
|
||
" <td>44.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>3</th>\n",
|
||
" <td>Rick and Morty</td>\n",
|
||
" <td>(2013– )</td>\n",
|
||
" <td>Animation, Adventure, Comedy</td>\n",
|
||
" <td>9.2</td>\n",
|
||
" <td>An animated series that follows the exploits o...</td>\n",
|
||
" <td>Stars:Justin Roiland, Chris Parnell, Spencer G...</td>\n",
|
||
" <td>414,849</td>\n",
|
||
" <td>23.0</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>4</th>\n",
|
||
" <td>Army of Thieves</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Crime, Horror</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>A prequel, set before the events of Army of th...</td>\n",
|
||
" <td>Director:Matthias Schweighöfer| Stars:Matt...</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR \\\n",
|
||
"0 Blood Red Sky (2021) \n",
|
||
"1 Masters of the Universe: Revelation (2021– ) \n",
|
||
"2 The Walking Dead (2010–2022) \n",
|
||
"3 Rick and Morty (2013– ) \n",
|
||
"4 Army of Thieves (2021) \n",
|
||
"\n",
|
||
" GENRE RATING \\\n",
|
||
"0 Action, Horror, Thriller 6.1 \n",
|
||
"1 Animation, Action, Adventure 5.0 \n",
|
||
"2 Drama, Horror, Thriller 8.2 \n",
|
||
"3 Animation, Adventure, Comedy 9.2 \n",
|
||
"4 Action, Crime, Horror NaN \n",
|
||
"\n",
|
||
" ONE-LINE \\\n",
|
||
"0 A woman with a mysterious illness is forced in... \n",
|
||
"1 The war for Eternia begins again in what may b... \n",
|
||
"2 Sheriff Deputy Rick Grimes wakes up from a com... \n",
|
||
"3 An animated series that follows the exploits o... \n",
|
||
"4 A prequel, set before the events of Army of th... \n",
|
||
"\n",
|
||
" STARS VOTES RunTime Gross \n",
|
||
"0 Director:Peter Thorwarth| Stars:Peri Baume... 21,062 121.0 NaN \n",
|
||
"1 Stars:Chris Wood, Sarah Michelle Gellar, Lena ... 17,870 25.0 NaN \n",
|
||
"2 Stars:Andrew Lincoln, Norman Reedus, Melissa M... 885,805 44.0 NaN \n",
|
||
"3 Stars:Justin Roiland, Chris Parnell, Spencer G... 414,849 23.0 NaN \n",
|
||
"4 Director:Matthias Schweighöfer| Stars:Matt... NaN NaN NaN "
|
||
]
|
||
},
|
||
"execution_count": 59,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"df['GENRE'] = df['GENRE'].str.replace('\\n','')\n",
|
||
"df['ONE-LINE'] = df['ONE-LINE'].str.replace('\\n','')\n",
|
||
"df['STARS'] = df['STARS'].str.replace('\\n','')\n",
|
||
"\n",
|
||
"df['GENRE'] = df['GENRE'].str.strip()\n",
|
||
"df['ONE-LINE'] = df['ONE-LINE'].str.strip()\n",
|
||
"df['STARS'] = df['STARS'].str.strip()\n",
|
||
"\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 60,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>Director:Peter Thorwarth| Stars:Peri Baume...</td>\n",
|
||
" <td>21,062</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>1</th>\n",
|
||
" <td>Masters of the Universe: Revelation</td>\n",
|
||
" <td>(2021– )</td>\n",
|
||
" <td>Animation, Action, Adventure</td>\n",
|
||
" <td>5.0</td>\n",
|
||
" <td>Stars:Chris Wood, Sarah Michelle Gellar, Lena ...</td>\n",
|
||
" <td>17,870</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>2</th>\n",
|
||
" <td>The Walking Dead</td>\n",
|
||
" <td>(2010–2022)</td>\n",
|
||
" <td>Drama, Horror, Thriller</td>\n",
|
||
" <td>8.2</td>\n",
|
||
" <td>Stars:Andrew Lincoln, Norman Reedus, Melissa M...</td>\n",
|
||
" <td>885,805</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>3</th>\n",
|
||
" <td>Rick and Morty</td>\n",
|
||
" <td>(2013– )</td>\n",
|
||
" <td>Animation, Adventure, Comedy</td>\n",
|
||
" <td>9.2</td>\n",
|
||
" <td>Stars:Justin Roiland, Chris Parnell, Spencer G...</td>\n",
|
||
" <td>414,849</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>4</th>\n",
|
||
" <td>Army of Thieves</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Crime, Horror</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" <td>Director:Matthias Schweighöfer| Stars:Matt...</td>\n",
|
||
" <td>NaN</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR \\\n",
|
||
"0 Blood Red Sky (2021) \n",
|
||
"1 Masters of the Universe: Revelation (2021– ) \n",
|
||
"2 The Walking Dead (2010–2022) \n",
|
||
"3 Rick and Morty (2013– ) \n",
|
||
"4 Army of Thieves (2021) \n",
|
||
"\n",
|
||
" GENRE RATING \\\n",
|
||
"0 Action, Horror, Thriller 6.1 \n",
|
||
"1 Animation, Action, Adventure 5.0 \n",
|
||
"2 Drama, Horror, Thriller 8.2 \n",
|
||
"3 Animation, Adventure, Comedy 9.2 \n",
|
||
"4 Action, Crime, Horror NaN \n",
|
||
"\n",
|
||
" STARS VOTES \n",
|
||
"0 Director:Peter Thorwarth| Stars:Peri Baume... 21,062 \n",
|
||
"1 Stars:Chris Wood, Sarah Michelle Gellar, Lena ... 17,870 \n",
|
||
"2 Stars:Andrew Lincoln, Norman Reedus, Melissa M... 885,805 \n",
|
||
"3 Stars:Justin Roiland, Chris Parnell, Spencer G... 414,849 \n",
|
||
"4 Director:Matthias Schweighöfer| Stars:Matt... NaN "
|
||
]
|
||
},
|
||
"execution_count": 60,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"del df['Gross'], df['RunTime'], df['ONE-LINE']\n",
|
||
"\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 61,
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"import re\n",
|
||
"\n",
|
||
"def get_start_year(year):\n",
|
||
" result = re.search(r'\\(([0-9]{4})\\-([0-9]{4})\\)',year)\n",
|
||
" if result:\n",
|
||
" return result.group(1)\n",
|
||
" result = re.search(r'\\(([0-9]{4})\\)',year)\n",
|
||
" if result:\n",
|
||
" return result.group(1)\n",
|
||
"\n",
|
||
"def get_end_year(year):\n",
|
||
" result = re.search(r'\\(([0-9]{4})\\-([0-9]{4})\\)',year)\n",
|
||
" if result:\n",
|
||
" return result.group(2)\n",
|
||
" result = re.search(r'\\(([0-9]{4})\\)',year)\n",
|
||
" if result:\n",
|
||
" return result.group(1)\n",
|
||
"\n",
|
||
"df.dropna(inplace=True)\n",
|
||
"df['start_year'] = df['YEAR'].apply(lambda y : get_start_year(y))\n",
|
||
"df['end_year'] = df['YEAR'].apply(lambda y : get_end_year(y))"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 62,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"0 Peri Baumeister, Carl Anton Koch, Alexander Sc...\n",
|
||
"6 Shailene Woodley, Joe Alwyn, Wendy Nottingham,...\n",
|
||
"10 Karen Gillan, Lena Headey, Carla Gugino, Miche...\n",
|
||
"12 Kiana Madeira, Olivia Scott Welch, Benjamin Fl...\n",
|
||
"20 Sadie Sink, Emily Rudd, Ryan Simpkins, McCabe ...\n",
|
||
"Name: stars, dtype: object"
|
||
]
|
||
},
|
||
"execution_count": 62,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"import re\n",
|
||
"\n",
|
||
"def extract_director(direct):\n",
|
||
" result = re.search(r'(Director:|Directors:)(.*)\\|',direct)\n",
|
||
" if result:\n",
|
||
" return result.group(2).strip()\n",
|
||
" return ''\n",
|
||
"\n",
|
||
"def extract_stars(stars):\n",
|
||
" result = re.search(r'(Stars:|Star:)(.*)',stars)\n",
|
||
" if result:\n",
|
||
" return result.group(2).strip()\n",
|
||
" return ''\n",
|
||
"\n",
|
||
"df['directors'] = df['STARS'].apply(lambda d : extract_director(d))\n",
|
||
"df['stars'] = df['STARS'].apply(lambda s : extract_stars(s))\n",
|
||
"df.dropna(inplace=True)\n",
|
||
"df['stars'].head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 63,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>start_year</th>\n",
|
||
" <th>end_year</th>\n",
|
||
" <th>directors</th>\n",
|
||
" <th>stars</th>\n",
|
||
" <th>genres</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>Director:Peter Thorwarth| Stars:Peri Baume...</td>\n",
|
||
" <td>21,062</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Peter Thorwarth</td>\n",
|
||
" <td>Peri Baumeister|Carl Anton Koch|Alexander Sche...</td>\n",
|
||
" <td>Action|Horror|Thriller</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>6</th>\n",
|
||
" <td>The Last Letter from Your Lover</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Romance</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Augustine Frizzell| Stars:Shailen...</td>\n",
|
||
" <td>5,283</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Augustine Frizzell</td>\n",
|
||
" <td>Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe...</td>\n",
|
||
" <td>Drama|Romance</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>10</th>\n",
|
||
" <td>Gunpowder Milkshake</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Adventure, Thriller</td>\n",
|
||
" <td>6.0</td>\n",
|
||
" <td>Director:Navot Papushado| Stars:Karen Gill...</td>\n",
|
||
" <td>17,989</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Navot Papushado</td>\n",
|
||
" <td>Karen Gillan|Lena Headey|Carla Gugino|Michelle...</td>\n",
|
||
" <td>Action|Adventure|Thriller</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>12</th>\n",
|
||
" <td>Fear Street: 1994</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.2</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Kiana Madeira...</td>\n",
|
||
" <td>50,148</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Kiana Madeira|Olivia Scott Welch|Benjamin Flor...</td>\n",
|
||
" <td>Drama|Horror|Mystery</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>20</th>\n",
|
||
" <td>Fear Street: 1978</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Sadie Sink, E...</td>\n",
|
||
" <td>36,634</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye</td>\n",
|
||
" <td>Drama|Horror|Mystery</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR GENRE \\\n",
|
||
"0 Blood Red Sky (2021) Action, Horror, Thriller \n",
|
||
"6 The Last Letter from Your Lover (2021) Drama, Romance \n",
|
||
"10 Gunpowder Milkshake (2021) Action, Adventure, Thriller \n",
|
||
"12 Fear Street: 1994 (2021) Drama, Horror, Mystery \n",
|
||
"20 Fear Street: 1978 (2021) Drama, Horror, Mystery \n",
|
||
"\n",
|
||
" RATING STARS VOTES \\\n",
|
||
"0 6.1 Director:Peter Thorwarth| Stars:Peri Baume... 21,062 \n",
|
||
"6 6.8 Director:Augustine Frizzell| Stars:Shailen... 5,283 \n",
|
||
"10 6.0 Director:Navot Papushado| Stars:Karen Gill... 17,989 \n",
|
||
"12 6.2 Director:Leigh Janiak| Stars:Kiana Madeira... 50,148 \n",
|
||
"20 6.8 Director:Leigh Janiak| Stars:Sadie Sink, E... 36,634 \n",
|
||
"\n",
|
||
" start_year end_year directors \\\n",
|
||
"0 2021 2021 Peter Thorwarth \n",
|
||
"6 2021 2021 Augustine Frizzell \n",
|
||
"10 2021 2021 Navot Papushado \n",
|
||
"12 2021 2021 Leigh Janiak \n",
|
||
"20 2021 2021 Leigh Janiak \n",
|
||
"\n",
|
||
" stars \\\n",
|
||
"0 Peri Baumeister|Carl Anton Koch|Alexander Sche... \n",
|
||
"6 Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe... \n",
|
||
"10 Karen Gillan|Lena Headey|Carla Gugino|Michelle... \n",
|
||
"12 Kiana Madeira|Olivia Scott Welch|Benjamin Flor... \n",
|
||
"20 Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye \n",
|
||
"\n",
|
||
" genres \n",
|
||
"0 Action|Horror|Thriller \n",
|
||
"6 Drama|Romance \n",
|
||
"10 Action|Adventure|Thriller \n",
|
||
"12 Drama|Horror|Mystery \n",
|
||
"20 Drama|Horror|Mystery "
|
||
]
|
||
},
|
||
"execution_count": 63,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"def strip_all_str(to_strip:str):\n",
|
||
" striped_str = []\n",
|
||
" for s in to_strip:\n",
|
||
" striped_str.append(s.strip())\n",
|
||
" return striped_str\n",
|
||
"\n",
|
||
"df['stars'] = df['stars'].str.split(',')\n",
|
||
"df['directors'] = df['directors'].str.split(',')\n",
|
||
"df['genres'] = df['GENRE'].str.split(',')\n",
|
||
"\n",
|
||
"df['stars'] = df['stars'].apply(lambda s : strip_all_str(s))\n",
|
||
"df['directors'] = df['directors'].apply(lambda s : strip_all_str(s))\n",
|
||
"df['genres'] = df['genres'].apply(lambda s : strip_all_str(s))\n",
|
||
"\n",
|
||
"df['stars'] = df['stars'].apply(lambda s : '|'.join(s))\n",
|
||
"df['directors'] = df['directors'].apply(lambda s : '|'.join(s))\n",
|
||
"df['genres'] = df['genres'].apply(lambda s : '|'.join(s))\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 64,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>start_year</th>\n",
|
||
" <th>end_year</th>\n",
|
||
" <th>directors</th>\n",
|
||
" <th>stars</th>\n",
|
||
" <th>...</th>\n",
|
||
" <th>Greg Kading</th>\n",
|
||
" <th>Griffin Gluck</th>\n",
|
||
" <th>Peri Baumeister</th>\n",
|
||
" <th>Greg Chun</th>\n",
|
||
" <th>Carlos Belloso</th>\n",
|
||
" <th>Carlos Barbosa</th>\n",
|
||
" <th>John Abraham</th>\n",
|
||
" <th>Mauricio Argüelles</th>\n",
|
||
" <th>Maurice Compte</th>\n",
|
||
" <th>John Belushi</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>Director:Peter Thorwarth| Stars:Peri Baume...</td>\n",
|
||
" <td>21,062</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Peter Thorwarth</td>\n",
|
||
" <td>Peri Baumeister|Carl Anton Koch|Alexander Sche...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>6</th>\n",
|
||
" <td>The Last Letter from Your Lover</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Romance</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Augustine Frizzell| Stars:Shailen...</td>\n",
|
||
" <td>5,283</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Augustine Frizzell</td>\n",
|
||
" <td>Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>10</th>\n",
|
||
" <td>Gunpowder Milkshake</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Adventure, Thriller</td>\n",
|
||
" <td>6.0</td>\n",
|
||
" <td>Director:Navot Papushado| Stars:Karen Gill...</td>\n",
|
||
" <td>17,989</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Navot Papushado</td>\n",
|
||
" <td>Karen Gillan|Lena Headey|Carla Gugino|Michelle...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>12</th>\n",
|
||
" <td>Fear Street: 1994</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.2</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Kiana Madeira...</td>\n",
|
||
" <td>50,148</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Kiana Madeira|Olivia Scott Welch|Benjamin Flor...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>20</th>\n",
|
||
" <td>Fear Street: 1978</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Sadie Sink, E...</td>\n",
|
||
" <td>36,634</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"<p>5 rows × 2038 columns</p>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR GENRE \\\n",
|
||
"0 Blood Red Sky (2021) Action, Horror, Thriller \n",
|
||
"6 The Last Letter from Your Lover (2021) Drama, Romance \n",
|
||
"10 Gunpowder Milkshake (2021) Action, Adventure, Thriller \n",
|
||
"12 Fear Street: 1994 (2021) Drama, Horror, Mystery \n",
|
||
"20 Fear Street: 1978 (2021) Drama, Horror, Mystery \n",
|
||
"\n",
|
||
" RATING STARS VOTES \\\n",
|
||
"0 6.1 Director:Peter Thorwarth| Stars:Peri Baume... 21,062 \n",
|
||
"6 6.8 Director:Augustine Frizzell| Stars:Shailen... 5,283 \n",
|
||
"10 6.0 Director:Navot Papushado| Stars:Karen Gill... 17,989 \n",
|
||
"12 6.2 Director:Leigh Janiak| Stars:Kiana Madeira... 50,148 \n",
|
||
"20 6.8 Director:Leigh Janiak| Stars:Sadie Sink, E... 36,634 \n",
|
||
"\n",
|
||
" start_year end_year directors \\\n",
|
||
"0 2021 2021 Peter Thorwarth \n",
|
||
"6 2021 2021 Augustine Frizzell \n",
|
||
"10 2021 2021 Navot Papushado \n",
|
||
"12 2021 2021 Leigh Janiak \n",
|
||
"20 2021 2021 Leigh Janiak \n",
|
||
"\n",
|
||
" stars ... Greg Kading \\\n",
|
||
"0 Peri Baumeister|Carl Anton Koch|Alexander Sche... ... 0 \n",
|
||
"6 Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe... ... 0 \n",
|
||
"10 Karen Gillan|Lena Headey|Carla Gugino|Michelle... ... 0 \n",
|
||
"12 Kiana Madeira|Olivia Scott Welch|Benjamin Flor... ... 0 \n",
|
||
"20 Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye ... 0 \n",
|
||
"\n",
|
||
" Griffin Gluck Peri Baumeister Greg Chun Carlos Belloso Carlos Barbosa \\\n",
|
||
"0 0 1 0 0 0 \n",
|
||
"6 0 0 0 0 0 \n",
|
||
"10 0 0 0 0 0 \n",
|
||
"12 0 0 0 0 0 \n",
|
||
"20 0 0 0 0 0 \n",
|
||
"\n",
|
||
" John Abraham Mauricio Argüelles Maurice Compte John Belushi \n",
|
||
"0 0 0 0 0 \n",
|
||
"6 0 0 0 0 \n",
|
||
"10 0 0 0 0 \n",
|
||
"12 0 0 0 0 \n",
|
||
"20 0 0 0 0 \n",
|
||
"\n",
|
||
"[5 rows x 2038 columns]"
|
||
]
|
||
},
|
||
"execution_count": 64,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"df_star_dummies = df['stars'].str.get_dummies(sep='|')\n",
|
||
"df_directors_dummies = df['directors'].str.get_dummies(sep='|')\n",
|
||
"df_genres_dummies = df['genres'].str.get_dummies(sep='|')\n",
|
||
"\n",
|
||
"most_important_values = list(df_star_dummies.sum().sort_values(ascending=False).head(1000).to_dict().keys())\n",
|
||
"df_star_dummies = df_star_dummies.loc[:,most_important_values]\n",
|
||
"most_important_values = list(df_directors_dummies.sum().sort_values(ascending=False).head(1000).to_dict().keys())\n",
|
||
"df_directors_dummies = df_directors_dummies.loc[:,most_important_values]\n",
|
||
"most_important_values = list(df_genres_dummies.sum().sort_values(ascending=False).head(1000).to_dict().keys())\n",
|
||
"df_genres_dummies = df_genres_dummies.loc[:,most_important_values]\n",
|
||
"\n",
|
||
"df = pd.concat([df,df_genres_dummies,df_directors_dummies,df_star_dummies],axis=1)\n",
|
||
"\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 65,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>MOVIES</th>\n",
|
||
" <th>YEAR</th>\n",
|
||
" <th>GENRE</th>\n",
|
||
" <th>RATING</th>\n",
|
||
" <th>STARS</th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>start_year</th>\n",
|
||
" <th>end_year</th>\n",
|
||
" <th>directors</th>\n",
|
||
" <th>stars</th>\n",
|
||
" <th>...</th>\n",
|
||
" <th>Greg Kading</th>\n",
|
||
" <th>Griffin Gluck</th>\n",
|
||
" <th>Peri Baumeister</th>\n",
|
||
" <th>Greg Chun</th>\n",
|
||
" <th>Carlos Belloso</th>\n",
|
||
" <th>Carlos Barbosa</th>\n",
|
||
" <th>John Abraham</th>\n",
|
||
" <th>Mauricio Argüelles</th>\n",
|
||
" <th>Maurice Compte</th>\n",
|
||
" <th>John Belushi</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>0</th>\n",
|
||
" <td>Blood Red Sky</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Horror, Thriller</td>\n",
|
||
" <td>6.1</td>\n",
|
||
" <td>Director:Peter Thorwarth| Stars:Peri Baume...</td>\n",
|
||
" <td>21062</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Peter Thorwarth</td>\n",
|
||
" <td>Peri Baumeister|Carl Anton Koch|Alexander Sche...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>6</th>\n",
|
||
" <td>The Last Letter from Your Lover</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Romance</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Augustine Frizzell| Stars:Shailen...</td>\n",
|
||
" <td>5283</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Augustine Frizzell</td>\n",
|
||
" <td>Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>10</th>\n",
|
||
" <td>Gunpowder Milkshake</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Action, Adventure, Thriller</td>\n",
|
||
" <td>6.0</td>\n",
|
||
" <td>Director:Navot Papushado| Stars:Karen Gill...</td>\n",
|
||
" <td>17989</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Navot Papushado</td>\n",
|
||
" <td>Karen Gillan|Lena Headey|Carla Gugino|Michelle...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>12</th>\n",
|
||
" <td>Fear Street: 1994</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.2</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Kiana Madeira...</td>\n",
|
||
" <td>50148</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Kiana Madeira|Olivia Scott Welch|Benjamin Flor...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>20</th>\n",
|
||
" <td>Fear Street: 1978</td>\n",
|
||
" <td>(2021)</td>\n",
|
||
" <td>Drama, Horror, Mystery</td>\n",
|
||
" <td>6.8</td>\n",
|
||
" <td>Director:Leigh Janiak| Stars:Sadie Sink, E...</td>\n",
|
||
" <td>36634</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>Leigh Janiak</td>\n",
|
||
" <td>Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"<p>5 rows × 2038 columns</p>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" MOVIES YEAR GENRE \\\n",
|
||
"0 Blood Red Sky (2021) Action, Horror, Thriller \n",
|
||
"6 The Last Letter from Your Lover (2021) Drama, Romance \n",
|
||
"10 Gunpowder Milkshake (2021) Action, Adventure, Thriller \n",
|
||
"12 Fear Street: 1994 (2021) Drama, Horror, Mystery \n",
|
||
"20 Fear Street: 1978 (2021) Drama, Horror, Mystery \n",
|
||
"\n",
|
||
" RATING STARS VOTES \\\n",
|
||
"0 6.1 Director:Peter Thorwarth| Stars:Peri Baume... 21062 \n",
|
||
"6 6.8 Director:Augustine Frizzell| Stars:Shailen... 5283 \n",
|
||
"10 6.0 Director:Navot Papushado| Stars:Karen Gill... 17989 \n",
|
||
"12 6.2 Director:Leigh Janiak| Stars:Kiana Madeira... 50148 \n",
|
||
"20 6.8 Director:Leigh Janiak| Stars:Sadie Sink, E... 36634 \n",
|
||
"\n",
|
||
" start_year end_year directors \\\n",
|
||
"0 2021 2021 Peter Thorwarth \n",
|
||
"6 2021 2021 Augustine Frizzell \n",
|
||
"10 2021 2021 Navot Papushado \n",
|
||
"12 2021 2021 Leigh Janiak \n",
|
||
"20 2021 2021 Leigh Janiak \n",
|
||
"\n",
|
||
" stars ... Greg Kading \\\n",
|
||
"0 Peri Baumeister|Carl Anton Koch|Alexander Sche... ... 0 \n",
|
||
"6 Shailene Woodley|Joe Alwyn|Wendy Nottingham|Fe... ... 0 \n",
|
||
"10 Karen Gillan|Lena Headey|Carla Gugino|Michelle... ... 0 \n",
|
||
"12 Kiana Madeira|Olivia Scott Welch|Benjamin Flor... ... 0 \n",
|
||
"20 Sadie Sink|Emily Rudd|Ryan Simpkins|McCabe Slye ... 0 \n",
|
||
"\n",
|
||
" Griffin Gluck Peri Baumeister Greg Chun Carlos Belloso Carlos Barbosa \\\n",
|
||
"0 0 1 0 0 0 \n",
|
||
"6 0 0 0 0 0 \n",
|
||
"10 0 0 0 0 0 \n",
|
||
"12 0 0 0 0 0 \n",
|
||
"20 0 0 0 0 0 \n",
|
||
"\n",
|
||
" John Abraham Mauricio Argüelles Maurice Compte John Belushi \n",
|
||
"0 0 0 0 0 \n",
|
||
"6 0 0 0 0 \n",
|
||
"10 0 0 0 0 \n",
|
||
"12 0 0 0 0 \n",
|
||
"20 0 0 0 0 \n",
|
||
"\n",
|
||
"[5 rows x 2038 columns]"
|
||
]
|
||
},
|
||
"execution_count": 65,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"df['start_year'] = df['start_year'].apply(lambda x : int(x))\n",
|
||
"df['end_year'] = df['end_year'].apply(lambda x : int(x))\n",
|
||
"df['RATING'] = df['RATING'].apply(lambda x : float(x))\n",
|
||
"\n",
|
||
"df['VOTES'] = df['VOTES'].str.replace(',','')\n",
|
||
"df['VOTES'] = df['VOTES'].apply(lambda x : int(x))\n",
|
||
"\n",
|
||
"df.head()"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 66,
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"cleaned_df = df\n",
|
||
"del cleaned_df['MOVIES'], cleaned_df['GENRE'],cleaned_df['STARS'],cleaned_df['directors'],cleaned_df['stars'],cleaned_df['genres'],cleaned_df['YEAR']\n",
|
||
"cleaned_df.dropna(inplace=True)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 67,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/html": [
|
||
"<div>\n",
|
||
"<style scoped>\n",
|
||
" .dataframe tbody tr th:only-of-type {\n",
|
||
" vertical-align: middle;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe tbody tr th {\n",
|
||
" vertical-align: top;\n",
|
||
" }\n",
|
||
"\n",
|
||
" .dataframe thead th {\n",
|
||
" text-align: right;\n",
|
||
" }\n",
|
||
"</style>\n",
|
||
"<table border=\"1\" class=\"dataframe\">\n",
|
||
" <thead>\n",
|
||
" <tr style=\"text-align: right;\">\n",
|
||
" <th></th>\n",
|
||
" <th>VOTES</th>\n",
|
||
" <th>start_year</th>\n",
|
||
" <th>end_year</th>\n",
|
||
" <th>Drama</th>\n",
|
||
" <th>Comedy</th>\n",
|
||
" <th>Documentary</th>\n",
|
||
" <th>Crime</th>\n",
|
||
" <th>Action</th>\n",
|
||
" <th>Thriller</th>\n",
|
||
" <th>Romance</th>\n",
|
||
" <th>...</th>\n",
|
||
" <th>Greg Kading</th>\n",
|
||
" <th>Griffin Gluck</th>\n",
|
||
" <th>Peri Baumeister</th>\n",
|
||
" <th>Greg Chun</th>\n",
|
||
" <th>Carlos Belloso</th>\n",
|
||
" <th>Carlos Barbosa</th>\n",
|
||
" <th>John Abraham</th>\n",
|
||
" <th>Mauricio Argüelles</th>\n",
|
||
" <th>Maurice Compte</th>\n",
|
||
" <th>John Belushi</th>\n",
|
||
" </tr>\n",
|
||
" </thead>\n",
|
||
" <tbody>\n",
|
||
" <tr>\n",
|
||
" <th>508</th>\n",
|
||
" <td>200206</td>\n",
|
||
" <td>2013</td>\n",
|
||
" <td>2013</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>509</th>\n",
|
||
" <td>34984</td>\n",
|
||
" <td>2018</td>\n",
|
||
" <td>2018</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>510</th>\n",
|
||
" <td>124972</td>\n",
|
||
" <td>1993</td>\n",
|
||
" <td>1993</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>512</th>\n",
|
||
" <td>21572</td>\n",
|
||
" <td>2018</td>\n",
|
||
" <td>2018</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>513</th>\n",
|
||
" <td>3082</td>\n",
|
||
" <td>2020</td>\n",
|
||
" <td>2020</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>...</th>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>...</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>9618</th>\n",
|
||
" <td>49</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>9814</th>\n",
|
||
" <td>175</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>9825</th>\n",
|
||
" <td>156</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>9826</th>\n",
|
||
" <td>145</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" <tr>\n",
|
||
" <th>9827</th>\n",
|
||
" <td>137</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>2021</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>1</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>...</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" <td>0</td>\n",
|
||
" </tr>\n",
|
||
" </tbody>\n",
|
||
"</table>\n",
|
||
"<p>3563 rows × 2030 columns</p>\n",
|
||
"</div>"
|
||
],
|
||
"text/plain": [
|
||
" VOTES start_year end_year Drama Comedy Documentary Crime Action \\\n",
|
||
"508 200206 2013 2013 0 1 0 0 0 \n",
|
||
"509 34984 2018 2018 1 0 0 1 0 \n",
|
||
"510 124972 1993 1993 1 0 0 0 0 \n",
|
||
"512 21572 2018 2018 1 0 0 1 0 \n",
|
||
"513 3082 2020 2020 1 0 0 0 0 \n",
|
||
"... ... ... ... ... ... ... ... ... \n",
|
||
"9618 49 2021 2021 0 1 0 0 0 \n",
|
||
"9814 175 2021 2021 0 0 1 1 0 \n",
|
||
"9825 156 2021 2021 0 0 1 1 0 \n",
|
||
"9826 145 2021 2021 0 0 1 1 0 \n",
|
||
"9827 137 2021 2021 0 0 1 1 0 \n",
|
||
"\n",
|
||
" Thriller Romance ... Greg Kading Griffin Gluck Peri Baumeister \\\n",
|
||
"508 0 0 ... 0 0 0 \n",
|
||
"509 0 0 ... 0 0 0 \n",
|
||
"510 1 0 ... 0 0 0 \n",
|
||
"512 0 0 ... 0 0 0 \n",
|
||
"513 0 0 ... 0 0 0 \n",
|
||
"... ... ... ... ... ... ... \n",
|
||
"9618 0 0 ... 0 0 0 \n",
|
||
"9814 0 0 ... 0 0 0 \n",
|
||
"9825 0 0 ... 0 0 0 \n",
|
||
"9826 0 0 ... 0 0 0 \n",
|
||
"9827 0 0 ... 0 0 0 \n",
|
||
"\n",
|
||
" Greg Chun Carlos Belloso Carlos Barbosa John Abraham \\\n",
|
||
"508 0 0 0 0 \n",
|
||
"509 0 0 0 0 \n",
|
||
"510 0 0 0 0 \n",
|
||
"512 0 0 0 0 \n",
|
||
"513 0 0 0 0 \n",
|
||
"... ... ... ... ... \n",
|
||
"9618 0 0 0 0 \n",
|
||
"9814 0 0 0 0 \n",
|
||
"9825 0 0 0 0 \n",
|
||
"9826 0 0 0 0 \n",
|
||
"9827 0 0 0 0 \n",
|
||
"\n",
|
||
" Mauricio Argüelles Maurice Compte John Belushi \n",
|
||
"508 0 0 0 \n",
|
||
"509 0 0 0 \n",
|
||
"510 0 0 0 \n",
|
||
"512 0 0 0 \n",
|
||
"513 0 0 0 \n",
|
||
"... ... ... ... \n",
|
||
"9618 0 0 0 \n",
|
||
"9814 0 0 0 \n",
|
||
"9825 0 0 0 \n",
|
||
"9826 0 0 0 \n",
|
||
"9827 0 0 0 \n",
|
||
"\n",
|
||
"[3563 rows x 2030 columns]"
|
||
]
|
||
},
|
||
"execution_count": 67,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"first_values = cleaned_df[200:]\n",
|
||
"features = first_values.loc[:, df.columns!='RATING']\n",
|
||
"labels = first_values[['RATING']].values\n",
|
||
"features"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 68,
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"from sklearn.model_selection import train_test_split\n",
|
||
"from sklearn.svm import SVC,LinearSVC\n",
|
||
"from sklearn.tree import DecisionTreeRegressor\n",
|
||
"from sklearn.ensemble import RandomForestRegressor\n",
|
||
"from sklearn.linear_model import LinearRegression\n",
|
||
"from sklearn.preprocessing import PolynomialFeatures\n",
|
||
"from sklearn.linear_model import Ridge\n",
|
||
"from sklearn.neighbors import KNeighborsClassifier\n",
|
||
"from sklearn.metrics import r2_score"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 69,
|
||
"metadata": {},
|
||
"outputs": [],
|
||
"source": [
|
||
"train, test, train_labels, test_labels = train_test_split(features,labels,test_size=0.33)"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 70,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"-0.03440859287218978"
|
||
]
|
||
},
|
||
"execution_count": 70,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"decision_tree = DecisionTreeRegressor()\n",
|
||
"decision_tree.fit(train, train_labels)\n",
|
||
"Y_pred = decision_tree.predict(test)\n",
|
||
"acc_decision_tree = r2_score(test_labels,Y_pred)\n",
|
||
"acc_decision_tree"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 71,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"-32814680366.75818"
|
||
]
|
||
},
|
||
"execution_count": 71,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"linreg = LinearRegression()\n",
|
||
"linreg.fit(train, train_labels)\n",
|
||
"Y_pred = linreg.predict(test)\n",
|
||
"acc_log = r2_score(test_labels,Y_pred)\n",
|
||
"acc_log"
|
||
]
|
||
},
|
||
{
|
||
"cell_type": "code",
|
||
"execution_count": 72,
|
||
"metadata": {},
|
||
"outputs": [
|
||
{
|
||
"data": {
|
||
"text/plain": [
|
||
"0.39096903616388345"
|
||
]
|
||
},
|
||
"execution_count": 72,
|
||
"metadata": {},
|
||
"output_type": "execute_result"
|
||
}
|
||
],
|
||
"source": [
|
||
"ridge_reg = Ridge(alpha=1,max_iter=100,tol=0.1)\n",
|
||
"ridge_reg.fit(train,train_labels)\n",
|
||
"Y_pred = ridge_reg.predict(test)\n",
|
||
"ridge_log = r2_score(test_labels,Y_pred)\n",
|
||
"ridge_log"
|
||
]
|
||
}
|
||
],
|
||
"metadata": {
|
||
"kernelspec": {
|
||
"display_name": "Python 3",
|
||
"language": "python",
|
||
"name": "python3"
|
||
},
|
||
"language_info": {
|
||
"codemirror_mode": {
|
||
"name": "ipython",
|
||
"version": 3
|
||
},
|
||
"file_extension": ".py",
|
||
"mimetype": "text/x-python",
|
||
"name": "python",
|
||
"nbconvert_exporter": "python",
|
||
"pygments_lexer": "ipython3",
|
||
"version": "3.8.10"
|
||
},
|
||
"orig_nbformat": 4,
|
||
"vscode": {
|
||
"interpreter": {
|
||
"hash": "916dbcbb3f70747c44a77c7bcd40155683ae19c65e1c03b4aa3499c5328201f1"
|
||
}
|
||
}
|
||
},
|
||
"nbformat": 4,
|
||
"nbformat_minor": 2
|
||
}
|