{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load in the Pitch() function from the mlpsoccer module to avoid having to build your own football pitch in a plot\n",
    "\n",
    "from mplsoccer.pitch import Pitch\n",
    "\n",
    "# pandas and numpy modules are loaded as pd and np to help with data handling\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Using pandas' read_csv() function you can read in the snippets of tracking data'\n",
    "# The function need you to write r'PATH TO YOUR DATA'\n",
    "\n",
    "home = pd.read_csv(r'C:\\\\Users\\\\YOUR_USER\\\\Desktop\\\\Github\\\\DFDA\\\\Data\\\\Tracking Data\\\\home.csv')\n",
    "\n",
    "away = pd.read_csv(r'C:\\\\Users\\\\YOUR_USER\\\\Desktop\\\\Github\\\\DFDA\\\\Data\\\\Tracking Data\\\\away.csv')\n",
    "\n",
    "info = pd.read_csv(r'C:\\\\Users\\\\YOUR_USER\\\\Desktop\\\\Github\\\\DFDA\\\\Data\\\\Tracking Data\\\\info.csv')\n",
    "\n",
    "# Again using head() you can take a look at the data\n",
    "home.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# You can select a series of frames by first using unique() and then [:x] to select x number of frames\n",
    "frames = info.frameIdx.unique()[:1]\n",
    "frames"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Subsetting the full dataset to only include the select frames is done by the isin() function\n",
    "\n",
    "homePlot = home[home['frameIdx'].isin(frames)]\n",
    "awayPlot = away[away['frameIdx'].isin(frames)]\n",
    "ballPlot = info[info['frameIdx'].isin(frames)]\n",
    "\n",
    "# To allow for an easier annotation of shirt numbers later you can extract the x, y, and shirtnumber column from each team\n",
    "\n",
    "homeNumbers=homePlot[[\"x\",\"y\",\"number\",\"team\"]]\n",
    "awayNumbers=awayPlot[[\"x\",\"y\",\"number\",\"team\"]]\n",
    "\n",
    "# And then combine them, one data set after the other by using pandas' concat() function\n",
    "\n",
    "numbers = pd.concat([homeNumbers, awayNumbers], ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# To understand how the code below adds shirt numbers to the plot - a breif description of loops are required\n",
    "# If you already understand loops just skip to the plot\n",
    "\n",
    "# You can make python go through certain things one step at a time by using loops\n",
    "# The following loop prints all the shirt numbers in the numbers data frame one at a time\n",
    "\n",
    "for i in range(0,len(numbers)):\n",
    "    print(numbers.team[i],numbers.number[i], \"- row\", i+1, \"done\") \n",
    "    # This prints the team, the shirt number and the row number+1 (as python starts counting from 0)\n",
    "    \n",
    "# It might go faster than you can notice, but the code does run one row at a time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# If you have chosen a single frame, the following will plot that frame\n",
    "# If you have chosen multiple frames, the plot will include all frames and be useless\n",
    "\n",
    "# First start by plotting a pitch, this time using second spectrum coordinates and setting the length and width of the pitch\n",
    "\n",
    "pitch = Pitch(pitch_type='secondspectrum', # Using second spectrum to fit with tracking data\n",
    "              pitch_width=68, pitch_length=105, # Setting pitch size to danish standard\n",
    "              goal_type='box', # Add \"nets\" to goals\n",
    "              pitch_color='grass', # Setting the color to be grass \n",
    "              line_color='white',  # Setting the lines on the pitch to be white\n",
    "              stripe=True, # Add stripes to the grass color\n",
    "              corner_arcs=True, # Add corner arcs\n",
    "             )\n",
    "\n",
    "fig, ax = pitch.draw(figsize=(16,10.4))\n",
    "\n",
    "# Then for each team plot the players by using a pitch.scatter() function\n",
    "\n",
    "scatter = pitch.scatter(homePlot.x, # x coordinate\n",
    "                        homePlot.y, # y coordinate\n",
    "                        ax=ax, # What plot to add them to \n",
    "                        edgecolor='black', # Edge color of the point\n",
    "                        c=homePlot.GK, # Color of the point\n",
    "                        s=250, # Size of the points\n",
    "                        cmap='autumn' # Color map - field players will be red and keeper yellow\n",
    "                       )\n",
    "\n",
    "scatter = pitch.scatter(awayPlot.x, # x coordinate\n",
    "                        awayPlot.y, # y coordinate\n",
    "                        ax=ax, # What plot to add them to \n",
    "                        edgecolor='black', # Edge color of the point\n",
    "                        c=awayPlot.GK, # Color of the point\n",
    "                        s=250, # Size of the points\n",
    "                        cmap='Set2' #Color map - field players will be green and keeper grey\n",
    "                       )\n",
    "\n",
    "\n",
    "# Add the ball first at a larger size, then a smaller size to create a point inside a point\n",
    "\n",
    "scatter = pitch.scatter(ballPlot.xBall, # x coordinate\n",
    "                        ballPlot.yBall, # y coordinate\n",
    "                        ax=ax, # What plot to add them to \n",
    "                        edgecolor='black', # Edge color of the point\n",
    "                        c='black', # Color of the point\n",
    "                        s=100 # Size of the points\n",
    "                       )\n",
    "\n",
    "scatter = pitch.scatter(ballPlot.xBall, # x coordinate\n",
    "                        ballPlot.yBall, # y coordinate\n",
    "                        ax=ax, # What plot to add them to \n",
    "                        edgecolor='black', # Edge color of the point\n",
    "                        c='white', # Color of the point\n",
    "                        s=40 # Size of the points\n",
    "                       )\n",
    "\n",
    "\n",
    "# Lastly loop over all the rows in the numbers dataframe and annotate each player with his shirt number\n",
    "\n",
    "for i in range(0,len(numbers)): # Loops from 0 to the number of rows - remember that python starts counting at 0\n",
    "    annotate = pitch.annotate(text=numbers.number[i], # Which text is printed - here the shirtnumber of row number i\n",
    "                             xy=(numbers.x[i],numbers.y[i]), # Where to print it - here x,y of row number i\n",
    "                             ax=ax, # What plot to add them to \n",
    "                             va='center', # Align the text at vertical center of x,y\n",
    "                             ha='center' # Align the text at horizontal center of x,y\n",
    "                             )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# To move from a single plot to an animation you need to choose a series of frames\n",
    "# The snippet variable has been created to split the data into segments each containing a goal - so you can subset on that\n",
    "# Or you can choose a series of frames like above - or subset based on the game clock by looking at info.gameClock\n",
    "\n",
    "# To plot it faster, you can plot every second frame - hence you can use [::2]\n",
    "\n",
    "frames = info.loc[info[\"snippet\"]==1].frameIdx.unique()[::2]\n",
    "\n",
    "df_ball=info[info['frameIdx'].isin(frames)]\n",
    "df_home=home[home['frameIdx'].isin(frames)]\n",
    "df_away=away[away['frameIdx'].isin(frames)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The function which creates and saves the plot are located in the Jupyter Notebook folder\n",
    "# To some extend it follows the ideas already used to plot a single frame\n",
    "\n",
    "from functions import save_match_clip\n",
    "\n",
    "# Saving the animation to your locals folder - might take a while depending on the number of frames!\n",
    "# Creators run time was 8+ minutes for all frames in snippet 1\n",
    "\n",
    "save_match_clip(df_ball, df_home, df_away, fpath=\"C:\\\\Users\\\\YOUR_USER\\\\Desktop\\\\Github\\\\DFDA\\\\Local\", fname='Clip', frames_per_second=12.5)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.3"
  },
  "vscode": {
   "interpreter": {
    "hash": "697d41c126625c3517a4890c41e0bb959a0eeddb5d944aeb5ca46094479b57e8"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}