intro_scikit_matplotlib.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction to `scikit` and `matplotlib`\n",
    "\n",
    "In this tutorial, we will use `numpy` to download global mean surface temperature data and then perform linear regression using `scikit` and plot the results with `matplotlib`, \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Starting again with our imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "Loading our data, that I have serialized from the arrays "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "xy = np.load(\"./data/temperature.npy\")"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Now your turn, `xy` has the first column corresponding to the year and the other to the temperature.\n",
    "Can you extract them into `x` and `y` respectively."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "x = ... # Your code\n",
    "\n",
    "y = ... # Your code\n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Sklearn\n",
    "Scikit-learn allows so simple machine learning (no neural network).  \n",
    "Here for simplicity, it will a Linear regression."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "model = LinearRegression()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "To use sklearn you need to define a model and then fit that model to the data, then after you can do prediction."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "# Reshape x to be a 2D array, necessary for scikit-learn\n",
    "x = x.reshape(-1, 1)\n",
    "\n",
    "model.fit(x, y)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Calculating the predictions"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "y_pred = model.predict(x)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Now what does this look like?\n",
    "Well let's plot it"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Matplotlib\n",
    "The main library to plot in python is matplotlib.\n",
    "It does get some time to get use to its quirks. (Do ask if you don't understand)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "I'll show you how to plot the x, y data and then your turn to plot the prediction on top of it, okay?"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "plt.scatter(x, y, label='Actual Data')\n",
    "plt.show()"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "Until you call `plt.show()` it will add the plots on the same image.\n",
    "Your turn to do plot *both* the actual data and the predicted one."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "outputs": [],
   "source": [
    "# Your code\n",
    "\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "You can also add the following for formatting:\n",
    "```python\n",
    "plt.xlabel('Year')\n",
    "plt.ylabel('Temperature')\n",
    "plt.title('Temperature vs Year')\n",
    "plt.legend()\n",
    "``` "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "An alternative library is called seaborn.\n",
    "You can ask me about it."
   ],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "language_info": {
   "name": "python"
  },
  "orig_nbformat": 4,
  "kernelspec": {
   "name": "python3",
   "language": "python",
   "display_name": "Python 3 (ipykernel)"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}