lectures/05_Data_Science.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Introduction to Python and Natural Language Technologies\n",
    "## Lecture 05, Data science, Handling text data, Pandas\n",
    "\n",
    "March 09, 2021\n",
    "\n",
    "Ádám Kovács\n",
    "\n",
    "The aim of the laboratory is to give a brief introduction to the popular Python library \"Pandas\". During the exercises we are going to use an IMDB dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "* Pandas is a popular Data science library for handling structural data. It has a lot of built in functions which you can explore [here](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.html)\n",
    "* Pandas is a great tool for doing Exploratory Data Analysis (EDA) and visualization on your data.\n",
    "* Before doing any Machine Learning, you always need to spend time to get to know your data.\n",
    "* Bash tools can be also used for quick data manipulations."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "First import the relevant libraries, you might have to install some of them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "!pip install pandas\n",
    "!pip install matplotlib\n",
    "!pip install numpy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Download and extract the data with Python code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "data_dir = os.getenv(\"data\")\n",
    "if data_dir is None:\n",
    "    data_dir = \"\"\n",
    "\n",
    "ml_path = os.path.join(data_dir, \"data.zip\")\n",
    "\n",
    "if not os.path.exists(ml_path):\n",
    "    print(\"Download data\")\n",
    "    import urllib\n",
    "    u = urllib.request.URLopener()\n",
    "    u.retrieve(\"http://sandbox.hlt.bme.hu/~adaamko/data.zip\", ml_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "Extracting:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "unzip_path = os.path.join(data_dir, \"data\")\n",
    "\n",
    "if not os.path.exists(unzip_path):\n",
    "    print(\"Unzip data\")\n",
    "    from zipfile import ZipFile\n",
    "    with ZipFile(ml_path) as myzip:\n",
    "        myzip.extractall(data_dir)\n",
    "        \n",
    "data_dir = unzip_path"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Get the occurence of each country from the release_dates file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "!cut -f3 -d, data/release_dates.csv | sort | uniq -c | sort"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### Count the most frequent words in the titles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "!cat data/titles.csv | sort | uniq -u | wc -l\n",
    "!cut -f1 -d, data/titles.csv | tr [:blank:] \\\\n | sort | uniq -c | sort | tail -10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "%matplotlib inline\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "\n",
    "matplotlib.style.use('ggplot')\n",
    "matplotlib.pyplot.rcParams['figure.figsize'] = (15, 3)\n",
    "matplotlib.pyplot.rcParams['font.family'] = 'sans-serif'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Loading datasets with pandas\n",
    "\n",
    "The function `pd.read_table` is used to load tabular data. It contains a lot of [parameters](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.read_table.html), with the file being the only required one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "%%time\n",
    "cast = pd.read_table('data/cast.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast = pd.read_table('data/cast.csv', sep=\",\", index_col=None)\n",
    "cast.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.tail(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.title\n",
    "cast[\"title\"]\n",
    "type(cast)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "`pd.read_table` function returns a `DataFrame` object. Columns can be selected as attributes or by their names between parentheses. If we select only one column it will return a `Series` object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "type(cast.title)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "More than one column can be also selected:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast[[\"title\", \"year\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "We can also select one column as a DataFrame, if we select that column as a list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "type(cast[[\"title\"]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "A DataFrame can be indexed in two ways, using either the `loc[]` or the `iloc[]` method. `iloc` returns a row by its position and accepts only integer value, `loc` returns a row by its label, which can be string also."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.iloc[1:6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.loc[1:6]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.loc[cast.year == 1985]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "We can also select columns with `iloc`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.iloc[:, 2:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Rows and columns can be retrieved in any order."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.iloc[[2,1,2], [1,1,3]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Selecting one cell:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.loc[0, \"name\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## First look on the data\n",
    "Let's look what information can be easily obtained from the DataFrame.\n",
    "\n",
    "Using the describe function we can acquire basic information about the rows: count, average, variance etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.year.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast.year.max()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "select only the movies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "titles = cast[['title', 'year']].drop_duplicates().reset_index(drop=True)\n",
    "titles.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles.sort_values(\"year\").tail(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Simple queries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# What movies were released in 1956?\n",
    "titles[titles.year == 1956]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# How many movies?\n",
    "len(titles[titles.year == 1956])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Querying the movies from the 90s\n",
    "\n",
    "titles[(titles.year >= 1990) & (titles.year <= 1999)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# List the Star Wars movies\n",
    "\n",
    "titles[titles.title == \"Star Wars\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles[titles.title.str.contains(\"Star Wars: \")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Movies that fall between A and B in the alphabet?\n",
    "titles[(titles.title >= 'A') & (titles.title <= 'B')]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "t = titles.copy()\n",
    "t = t.set_index('title').sort_index()\n",
    "t.loc['A':'B']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Oldest and newest movies\n",
    "titles.iloc[titles.year.idxmax()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Longest title\n",
    "t = titles.copy()\n",
    "t.iloc[titles.title.str.len().idxmax()]\n",
    "#t[t.title.str.match(\"^The .*\")].sort_values(by=\"year\")\n",
    "#t[t.title.str.match(r\"^[a-zA-Z]+\\s[a-zA-Z]+$\")]\n",
    "#t.title.str.replace(r\"^[a-zA-Z]+\\s[a-zA-Z]+$\", \"two words\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles.iloc[titles.title.str.len().idxmin()]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Grouping and visualization\n",
    "How many movies were released by year?\n",
    "\n",
    "We can answer the question in two steps:\n",
    "\n",
    "* grouping by years\n",
    "* aggregate for each year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles.groupby('year').size().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles.groupby('year').size().plot(kind=\"bar\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "titles[(titles.year > 1999) & (titles.year < 2010)].groupby('year').size().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast[cast.name == \"Matthew McConaughey\"].groupby(\"year\").size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Show a stacked diagram about the \"n\"(role number) values in the 90s\n",
    "i = cast.set_index('name').sort_index()\n",
    "i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "i.loc[\"Anthony Hopkins\", [\"year\", \"n\"]].groupby(\"year\").size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "i.loc[\"Anthony Hopkins\", [\"year\", \"n\"]].groupby(\"year\").agg([\"min\", \"mean\", \"max\"]).plot(kind=\"area\", stacked=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "Operators are overloaded so expressions are done on every row"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast.year + 4000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "t = titles.copy()\n",
    "t[\"movie_token_len\"] = t.title.apply(lambda x: len(x.split()))\n",
    "t.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "t.groupby(\"year\").mean().plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# What are the 11 most common character names in movie history?\n",
    "\n",
    "cast.character.value_counts().head(11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Which ten people have appeared most often as “Himself” over the history of film?\n",
    "\n",
    "c = cast.copy()\n",
    "c = c[c.character == 'Himself']\n",
    "c = c.groupby('name').size()\n",
    "c.sort_values().tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "c = cast.groupby([\"character\", \"type\"]).size()\n",
    "c"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### How many actors are in the cast of each version of Hamlet?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "cast[cast.title.str.startswith(\"Hamlet\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_by_title_year = cast.set_index(['title', 'year']).sort_index()\n",
    "cast_by_title_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_by_title_year.loc[\"Hamlet\", 1910]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# How many actors are in the cast of each version of Hamlet?\n",
    "\n",
    "cast_by_title_year.loc[\"Hamlet\": \"Hamlet (Z\"]\n",
    "#cast_by_title_year.loc[\"Hamlet\": \"Hamlet (Z\"].groupby(level=[0,1]).size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "#    Build a dataframe with a row for each year with two columns:\n",
    "#    the number of roles for actors in that year's films,\n",
    "#    and the number of roles for actresses.\n",
    "\n",
    "aa = cast[['year', 'type']].groupby(['year', 'type']).size()\n",
    "aa.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "aa.unstack(level=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "aa = aa.unstack()\n",
    "aa = aa.loc[1990:2010]\n",
    "aa"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "The `stack` and `unstack` functions are used to reshape the data into a stacked/unstacked format. \n",
    "![stack](https://www.w3resource.com/w3r_images/pandas-dataframe-unstack-1.svg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "aa.plot(kind='area')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## Release dates and merging"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "release_dates = pd.read_csv('data/release_dates.csv', index_col=None,\n",
    "                            parse_dates=['date'], infer_datetime_format=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "release_dates.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "release_dates.groupby(release_dates.date.dt.month).size().plot()\n",
    "#release_dates.groupby(release_dates.date.dt.day).size().plot(kind=\"pie\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "**We could ask the question of when X actors movies were released?**. \n",
    "For that we need to merge a subset of the cast table to the release_dates table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "rd = release_dates.set_index(\"title\").sort_index()\n",
    "rd.drop_duplicates()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_leonardo = cast[cast.name == \"Leonardo DiCaprio\"]\n",
    "cast_leonardo.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_date_leo = pd.merge(cast_leonardo, rd, left_on=\"title\", right_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_date_leo.groupby(\"country\").size().idxmax()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "cast_date_leo.groupby(cast_date_leo.date.dt.month).size().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "\"Pivot\": Reshape data (produce a “pivot” table) based on column values. Uses unique values from specified index / columns to form axes of the resulting DataFrame. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "r = release_dates\n",
    "r = r[r.title.str.startswith('Star Wars: Episode')]\n",
    "r = r[r.country.str.startswith('H')]\n",
    "r.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "piv = r.pivot('title', 'country', 'date')\n",
    "#piv = r.pivot('country', 'title', 'date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "piv.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "#### Pandas has GUI support  (see https://pypi.org/project/pandasgui/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "!pip install pandasgui"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": [
    "from pandasgui import show\n",
    "\n",
    "gui = show(titles.title)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# Exercises?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 1. Display the cast of \"The Lord of the Rings: The Fellowship of the Ring\" in their correct \"n\"-value order, ignoring roles that did not earn a numeric \"n\" value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 2. What are the 10 most frequent roles that contains the word \"teacher\"?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 3. List each of the characters that Jackie Chan has portrayed at least twice. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 4. Plot the difference between the number of actor roles each year and the number of actress roles each year over the history of film."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "## 5. Make a bar plot of the day of the week on which movies with \"Romance\" in their title tend to be released in the USA."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}