03_tutorial_NetworkTrainingKeras.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizing the training of Neural Networks with Keras"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Example code for the lecture series \"Machine Learning for Physicists\" by Florian Marquardt\n",
    "\n",
    "Lecture 3, Tutorials\n",
    "\n",
    "See https://machine-learning-for-physicists.org and the current course website linked there!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This notebook shows how to:\n",
    "- visualize the training of neural networks using keras\n",
    "\n",
    "The networks have 2 input and 1 output neurons, but arbitrarily many hidden layers, and also you can choose the activation functions\n",
    "\n",
    "This is essentially an extension of the lecture-2 training notebook, but now using keras instead of pure python."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Imports: numpy and matplotlib and keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# keras: Sequential is the neural-network class, Dense is\n",
    "# the standard network layer\n",
    "from tensorflow.keras import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras import optimizers # to choose more advanced optimizers like 'adam'\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt # for plotting\n",
    "import matplotlib\n",
    "matplotlib.rcParams['figure.dpi']=300 # highres display\n",
    "\n",
    "# for subplots within subplots:\n",
    "from matplotlib import gridspec\n",
    "\n",
    "# for nice inset colorbars: (approach changed from lecture 1 'Visualization' notebook)\n",
    "from mpl_toolkits.axes_grid1.inset_locator import InsetPosition\n",
    "\n",
    "# for updating display \n",
    "# (very simple animation)\n",
    "from IPython.display import clear_output\n",
    "from time import sleep"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# backpropagation and training routines\n",
    "# these are now just front-ends to keras!\n",
    "\n",
    "def apply_net(y_in): # one forward pass through the network\n",
    "    global Net    \n",
    "    return(Net.predict_on_batch(y_in))\n",
    "\n",
    "def apply_net_simple(y_in): # one forward pass through the network\n",
    "    return(apply_net(y_in))\n",
    "        \n",
    "def train_net(y_in,y_target): # one full training batch\n",
    "    # y_in is an array of size batchsize x (input-layer-size)\n",
    "    # y_target is an array of size batchsize x (output-layer-size)\n",
    "    global Net\n",
    "    \n",
    "    cost=Net.train_on_batch(y_in,y_target)[0]\n",
    "    return(cost)\n",
    "\n",
    "def init_layer_variables(weights,biases,activations,use_keras_init=False,eta=0.1,optimizer='sgd'):\n",
    "    global Weights, Biases, NumLayers, Activations\n",
    "    global LayerSizes, y_layer, df_layer, dw_layer, db_layer\n",
    "    global Net\n",
    "    \n",
    "    # store the main data in global variables\n",
    "    Weights=weights\n",
    "    Biases=biases\n",
    "    Activations=activations\n",
    "    NumLayers=len(Weights)\n",
    "\n",
    "    # keras activation names can be slightly different from what I used...\n",
    "    # also: 'jump' is not implemented here\n",
    "    KerasActivation={ \"sigmoid\":\"sigmoid\", \"reLU\":\"relu\", \"linear\":\"linear\" }\n",
    "\n",
    "    Net=Sequential() # a new network ('sequential' is the simplest structure: layer by layer)\n",
    "    \n",
    "    # now build up the network, layer by layer:\n",
    "    LayerSizes=[2]\n",
    "    for j in range(NumLayers):\n",
    "        LayerSizes.append(len(Biases[j]))\n",
    "        if use_keras_init: # use keras' random weight initialization approach\n",
    "            Net.add(Dense(len(Biases[j]), # number of neurons\n",
    "                          input_shape=(LayerSizes[j],), # size of previous layer\n",
    "                          activation=KerasActivation[activations[j]] # activation function\n",
    "                     ))            \n",
    "        else:\n",
    "            Net.add(Dense(len(Biases[j]), # number of neurons\n",
    "                          input_shape=(LayerSizes[j],), # size of previous layer\n",
    "                          activation=KerasActivation[activations[j]], # activation function\n",
    "                         weights = [ np.array(weights[j]), np.array(biases[j]) ] # the weights and biases for this layer\n",
    "                     ))\n",
    "    if optimizer=='adam':\n",
    "        the_optimizer=optimizers.Adam(lr=eta) # adaptive\n",
    "    else:\n",
    "        the_optimizer=optimizers.SGD(lr=eta) # standard gradient descent with given learning rate!\n",
    "    Net.compile(loss='mean_squared_error',\n",
    "              optimizer=the_optimizer,\n",
    "              metrics=['accuracy'])\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# visualization routines:\n",
    "\n",
    "# some internal routines for plotting the network:\n",
    "def plot_connection_line(ax,X,Y,W,vmax=1.0,linewidth=3):\n",
    "    t=np.linspace(0,1,20)\n",
    "    if W>0:\n",
    "        col=[0,0.4,0.8]\n",
    "    else:\n",
    "        col=[1,0.3,0]\n",
    "    ax.plot(X[0]+(3*t**2-2*t**3)*(X[1]-X[0]),Y[0]+t*(Y[1]-Y[0]),\n",
    "           alpha=abs(W)/vmax,color=col,\n",
    "           linewidth=linewidth)\n",
    "    \n",
    "def plot_neuron_alpha(ax,X,Y,B,size=100.0,vmax=1.0):\n",
    "    if B>0:\n",
    "        col=np.array([[0,0.4,0.8]])\n",
    "    else:\n",
    "        col=np.array([[1,0.3,0]])\n",
    "    ax.scatter([X],[Y],marker='o',c=col,alpha=abs(B)/vmax,s=size,zorder=10)\n",
    "\n",
    "def plot_neuron(ax,X,Y,B,size=100.0,vmax=1.0):\n",
    "    if B>0:\n",
    "        col=np.array([[0,0.4,0.8]])\n",
    "    else:\n",
    "        col=np.array([[1,0.3,0]])\n",
    "    ax.scatter([X],[Y],marker='o',c=col,s=size,zorder=10)\n",
    "    \n",
    "def visualize_network(weights,biases,activations,\n",
    "                      M=100,y0range=[-1,1],y1range=[-1,1],\n",
    "                     size=400.0, linewidth=5.0,\n",
    "                     weights_are_swapped=False,\n",
    "                    layers_already_initialized=False,\n",
    "                      plot_cost_function=None,\n",
    "                      current_cost=None, cost_max=None, plot_target=None\n",
    "                     ):\n",
    "    \"\"\"\n",
    "    Visualize a neural network with 2 input \n",
    "    neurons and 1 output neuron (plot output vs input in a 2D plot)\n",
    "    \n",
    "    weights is a list of the weight matrices for the\n",
    "    layers, where weights[j] is the matrix for the connections\n",
    "    from layer j to layer j+1 (where j==0 is the input)\n",
    "    \n",
    "    weights[j][m,k] is the weight for input neuron k going to output neuron m\n",
    "    (note: internally, m and k are swapped, see the explanation of\n",
    "    batch processing in lecture 2)\n",
    "    \n",
    "    biases[j] is the vector of bias values for obtaining the neurons in layer j+1\n",
    "    biases[j][k] is the bias for neuron k in layer j+1\n",
    "    \n",
    "    activations is a list of the activation functions for\n",
    "    the different layers: choose 'linear','sigmoid',\n",
    "    'jump' (i.e. step-function), and 'reLU'\n",
    "    \n",
    "    M is the resolution (MxM grid)\n",
    "    \n",
    "    y0range is the range of y0 neuron values (horizontal axis)\n",
    "    y1range is the range of y1 neuron values (vertical axis)\n",
    "    \"\"\"\n",
    "    if not weights_are_swapped:\n",
    "        swapped_weights=[]\n",
    "        for j in range(len(weights)):\n",
    "            swapped_weights.append(np.transpose(weights[j]))\n",
    "    else:\n",
    "        swapped_weights=weights\n",
    "\n",
    "    y0,y1=np.meshgrid(np.linspace(y0range[0],y0range[1],M),np.linspace(y1range[0],y1range[1],M))\n",
    "    y_in=np.zeros([M*M,2])\n",
    "    y_in[:,0]=y0.flatten()\n",
    "    y_in[:,1]=y1.flatten()\n",
    "    \n",
    "    # if we call visualization directly, we still\n",
    "    # need to initialize the 'Weights' and other\n",
    "    # global variables; otherwise (during training)\n",
    "    # all of this has already been taken care of:\n",
    "    if not layers_already_initialized:\n",
    "        init_layer_variables(swapped_weights,biases,activations)\n",
    "    y_out=apply_net_simple(y_in)\n",
    "\n",
    "    if plot_cost_function is None:\n",
    "        fig,ax=plt.subplots(ncols=2,nrows=1,figsize=(8,4))\n",
    "    else:\n",
    "        fig=plt.figure(figsize=(8,4))\n",
    "        gs_top = gridspec.GridSpec(nrows=1, ncols=2)\n",
    "        gs_left = gridspec.GridSpecFromSubplotSpec(nrows=2, ncols=1, subplot_spec=gs_top[0], height_ratios=[1.0,0.3])\n",
    "        ax=[ fig.add_subplot(gs_left[0]),\n",
    "            fig.add_subplot(gs_top[1]),\n",
    "           fig.add_subplot(gs_left[1]) ]\n",
    "        # ax[0] is network\n",
    "        # ax[1] is image produced by network\n",
    "        # ax[2] is cost function subplot\n",
    "        \n",
    "    # plot the network itself:\n",
    "    \n",
    "    # positions of neurons on plot:\n",
    "    posX=[[-0.5,+0.5]]; posY=[[0,0]]\n",
    "    vmax=0.0 # for finding the maximum weight\n",
    "    vmaxB=0.0 # for maximum bias\n",
    "    for j in range(len(biases)):\n",
    "        n_neurons=len(biases[j])\n",
    "        posX.append(np.array(range(n_neurons))-0.5*(n_neurons-1))\n",
    "        posY.append(np.full(n_neurons,j+1))\n",
    "        vmax=np.maximum(vmax,np.max(np.abs(weights[j])))\n",
    "        vmaxB=np.maximum(vmaxB,np.max(np.abs(biases[j])))\n",
    "\n",
    "    # plot connections\n",
    "    for j in range(len(biases)):\n",
    "        for k in range(len(posX[j])):\n",
    "            for m in range(len(posX[j+1])):\n",
    "                plot_connection_line(ax[0],[posX[j][k],posX[j+1][m]],\n",
    "                                     [posY[j][k],posY[j+1][m]],\n",
    "                                     swapped_weights[j][k,m],vmax=vmax,\n",
    "                                    linewidth=linewidth)\n",
    "    \n",
    "    # plot neurons\n",
    "    for k in range(len(posX[0])): # input neurons (have no bias!)\n",
    "        plot_neuron(ax[0],posX[0][k],posY[0][k],\n",
    "                   vmaxB,vmax=vmaxB,size=size)\n",
    "    for j in range(len(biases)): # all other neurons\n",
    "        for k in range(len(posX[j+1])):\n",
    "            plot_neuron(ax[0],posX[j+1][k],posY[j+1][k],\n",
    "                       biases[j][k],vmax=vmaxB,size=size)\n",
    "    \n",
    "    ax[0].axis('off')\n",
    "    \n",
    "    # now: the output of the network\n",
    "    img=ax[1].imshow(np.reshape(y_out,[M,M]),origin='lower',\n",
    "                    extent=[y0range[0],y0range[1],y1range[0],y1range[1]])\n",
    "    ax[1].set_xlabel(r'$y_0$')\n",
    "    ax[1].set_ylabel(r'$y_1$')\n",
    "    \n",
    "#     axins1 = inset_axes(ax[1],\n",
    "#                     width=\"40%\",  # width = 50% of parent_bbox width\n",
    "#                     height=\"5%\",  # height : 5%\n",
    "#                     loc='upper right',\n",
    "#                        bbox_to_anchor=[0.3,0.4])\n",
    "\n",
    "#    axins1 = ax[1].inset_axes([0.5,0.8,0.45,0.1])\n",
    "    axins1 = plt.axes([0, 0, 1, 1])\n",
    "    ip = InsetPosition(ax[1], [0.25, 0.1, 0.5, 0.05])\n",
    "    axins1.set_axes_locator(ip)\n",
    "\n",
    "    imgmin=np.min(y_out)\n",
    "    imgmax=np.max(y_out)\n",
    "    color_bar=fig.colorbar(img, cax=axins1, orientation=\"horizontal\",ticks=np.linspace(imgmin,imgmax,3))\n",
    "    cbxtick_obj = plt.getp(color_bar.ax.axes, 'xticklabels')\n",
    "    plt.setp(cbxtick_obj, color=\"white\")\n",
    "    axins1.xaxis.set_ticks_position(\"bottom\")\n",
    "\n",
    "    if plot_target is not None:\n",
    "        axins2 = plt.axes([0.01, 0.01, 0.99, 0.99])\n",
    "        ip = InsetPosition(ax[1], [0.75, 0.75, 0.2, 0.2])\n",
    "        axins2.set_axes_locator(ip)\n",
    "        axins2.imshow(plot_target,origin='lower')\n",
    "        axins2.get_xaxis().set_ticks([])\n",
    "        axins2.get_yaxis().set_ticks([])\n",
    "        \n",
    "    if plot_cost_function is not None:\n",
    "        ax[2].plot(plot_cost_function)\n",
    "        ax[2].set_ylim([0.0,cost_max])\n",
    "        ax[2].set_yticks([0.0,cost_max])\n",
    "        ax[2].set_yticklabels([\"0\",'{:1.2e}'.format(cost_max)])\n",
    "        if current_cost is not None:\n",
    "            ax[2].text(0.9, 0.9, 'cost={:1.2e}'.format(current_cost), horizontalalignment='right',\n",
    "                       verticalalignment='top', transform=ax[2].transAxes)\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "def visualize_network_training(weights,biases,activations,\n",
    "                               target_function,\n",
    "                               num_neurons=None,\n",
    "                               weight_scale=1.0,\n",
    "                               bias_scale=1.0,\n",
    "                               yspread=1.0,\n",
    "                      M=100,y0range=[-1,1],y1range=[-1,1],\n",
    "                     size=400.0, linewidth=5.0,\n",
    "                    steps=100, batchsize=10, eta=0.1,\n",
    "                              random_init=False,\n",
    "                              visualize_nsteps=1,\n",
    "                              plot_target=True,\n",
    "                              use_keras_init=False,\n",
    "                              optimizer='sgd'):\n",
    "    \"\"\"\n",
    "    Visualize the training of a neural network.\n",
    "    \n",
    "    weights, biases, and activations define the neural network \n",
    "    (the starting point of the optimization; for the detailed description,\n",
    "    see the help for visualize_network)\n",
    "    \n",
    "    If you want to have layers randomly initialized, just provide\n",
    "    the number of neurons for each layer as 'num_neurons'. This should include\n",
    "    all layers, including input (2 neurons) and output (1), so num_neurons=[2,3,5,4,1] is\n",
    "    a valid example. In this case, weight_scale and bias_scale define the\n",
    "    spread of the random Gaussian variables used to initialize all weights and biases.\n",
    "    \n",
    "    target_function is the name of the function that we\n",
    "    want to approximate; it must be possible to \n",
    "    evaluate this function on a batch of samples, by\n",
    "    calling target_function(y) on an array y of \n",
    "    shape [batchsize,2], where\n",
    "    the second index refers to the two coordinates\n",
    "    (input neuron values) y0 and y1. The return\n",
    "    value must be an array with one index, corresponding\n",
    "    to the batchsize. A valid example is:\n",
    "    \n",
    "    def my_target(y):\n",
    "        return( np.sin(y[:,0]) + np.cos(y[:,1]) )\n",
    "    \n",
    "    steps is the number of training steps\n",
    "    \n",
    "    batchsize is the number of samples per training step\n",
    "    \n",
    "    eta is the learning rate (stepsize in the gradient descent)\n",
    "    \n",
    "    yspread denotes the spread of the Gaussian\n",
    "    used to sample points in (y0,y1)-space\n",
    "    \n",
    "    visualize_n_steps>1 means skip some steps before\n",
    "    visualizing again (can speed up things)\n",
    "    \n",
    "    plot_target=True means do plot the target function in a corner\n",
    "    \n",
    "    For all the other parameters, see the help for\n",
    "        visualize_network\n",
    "    \n",
    "    weights and biases as given here will be used\n",
    "    as starting points, unless you specify\n",
    "    random_init=True, in which case they will be\n",
    "    used to determine the spread of Gaussian random\n",
    "    variables used for initialization!\n",
    "    \"\"\"\n",
    "    global Net, Weights, Biases\n",
    "    \n",
    "    if num_neurons is not None: # build weight matrices as randomly initialized\n",
    "        weights=[weight_scale*np.random.randn(num_neurons[j+1],num_neurons[j]) for j in range(len(num_neurons)-1)]\n",
    "        biases=[bias_scale*np.random.randn(num_neurons[j+1]) for j in range(len(num_neurons)-1)]\n",
    "    \n",
    "    swapped_weights=[]\n",
    "    for j in range(len(weights)):\n",
    "        swapped_weights.append(np.transpose(weights[j]))\n",
    "    init_layer_variables(swapped_weights,biases,activations,use_keras_init=use_keras_init,eta=eta,optimizer=optimizer)\n",
    "    \n",
    "    if plot_target:\n",
    "        y0,y1=np.meshgrid(np.linspace(y0range[0],y0range[1],M),np.linspace(y1range[0],y1range[1],M))\n",
    "        y=np.zeros([M*M,2])\n",
    "        y[:,0]=y0.flatten()\n",
    "        y[:,1]=y1.flatten()\n",
    "        plot_target_values=np.reshape(target_function(y),[M,M])\n",
    "    else:\n",
    "        plot_target_values=None\n",
    "    \n",
    "    y_target=np.zeros([batchsize,1])\n",
    "    costs=np.zeros(steps)\n",
    "    \n",
    "    for j in range(steps):\n",
    "        # produce samples (random points in y0,y1-space):\n",
    "        y_in=yspread*np.random.randn(batchsize,2)\n",
    "        # apply target function to those points:\n",
    "        y_target[:,0]=target_function(y_in)\n",
    "        # do one training step on this batch of samples:\n",
    "        costs[j]=train_net(y_in,y_target)\n",
    "        \n",
    "        # now visualize the updated network:\n",
    "        if j%visualize_nsteps==0:\n",
    "            clear_output(wait=True) # for animation\n",
    "            if j>10:\n",
    "                cost_max=np.average(costs[0:j])*1.5\n",
    "            else:\n",
    "                cost_max=costs[0]\n",
    "\n",
    "            # extract weights and biases from the keras network:\n",
    "            for j in range(NumLayers):\n",
    "                Weights[j],Biases[j]=Net.layers[j].get_weights()\n",
    "            \n",
    "            visualize_network(Weights,Biases,activations,\n",
    "                          M,y0range=y0range,y1range=y1range,\n",
    "                         size=size, linewidth=linewidth,\n",
    "                             weights_are_swapped=True,\n",
    "                             layers_already_initialized=True,\n",
    "                             plot_cost_function=costs,\n",
    "                             current_cost=costs[j],\n",
    "                             cost_max=cost_max,\n",
    "                             plot_target=plot_target_values)\n",
    "            sleep(0.1) # wait a bit before next step (probably not needed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Example 1: Training for a simple AND function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAACGQAAAQpCAYAAACNqEmwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBgAAAAAAAADAyAQyAAAAAAAAAABG9v8gTSkHfzSF/wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x13cf985c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def my_target(y):\n",
    "    return( 1.0*( (y[:,0]+y[:,1])>0) )\n",
    "\n",
    "visualize_network_training(weights=[ [ \n",
    "    [0.2,-0.9]  # weights of 2 input neurons for single output neuron\n",
    "    ] ],\n",
    "    biases=[ \n",
    "        [0.0] # bias for single output neuron\n",
    "            ],\n",
    "    target_function=my_target, # the target function to approximate\n",
    "    activations=[ 'sigmoid' # activation for output\n",
    "                ],\n",
    "    y0range=[-3,3],y1range=[-3,3],\n",
    "    steps=1000, eta=.5, batchsize=200,\n",
    "                          visualize_nsteps=10, \n",
    "                           plot_target=True,\n",
    "                          optimizer='adam')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Example 2: Training for half a smiley (circle with two eyes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-67-f51a23aa19d1>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     12\u001b[0m                            \u001b[0mplot_target\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     13\u001b[0m                           \u001b[0moptimizer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'adam'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 14\u001b[0;31m                           size=20,linewidth=2)\n\u001b[0m",
      "\u001b[0;32m<ipython-input-55-ccc9db246a6b>\u001b[0m in \u001b[0;36mvisualize_network_training\u001b[0;34m(weights, biases, activations, target_function, num_neurons, weight_scale, bias_scale, yspread, M, y0range, y1range, size, linewidth, steps, batchsize, eta, random_init, visualize_nsteps, plot_target, use_keras_init, optimizer)\u001b[0m\n\u001b[1;32m    281\u001b[0m                              \u001b[0mcurrent_cost\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcosts\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    282\u001b[0m                              \u001b[0mcost_max\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcost_max\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 283\u001b[0;31m                              plot_target=plot_target_values)\n\u001b[0m\u001b[1;32m    284\u001b[0m             \u001b[0msleep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0.1\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# wait a bit before next step (probably not needed)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-55-ccc9db246a6b>\u001b[0m in \u001b[0;36mvisualize_network\u001b[0;34m(weights, biases, activations, M, y0range, y1range, size, linewidth, weights_are_swapped, layers_already_initialized, plot_cost_function, current_cost, cost_max, plot_target)\u001b[0m\n\u001b[1;32m    120\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mk\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mposX\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    121\u001b[0m             plot_neuron(ax[0],posX[j+1][k],posY[j+1][k],\n\u001b[0;32m--> 122\u001b[0;31m                        biases[j][k],vmax=vmaxB,size=size)\n\u001b[0m\u001b[1;32m    123\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    124\u001b[0m     \u001b[0max\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'off'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-55-ccc9db246a6b>\u001b[0m in \u001b[0;36mplot_neuron\u001b[0;34m(ax, X, Y, B, size, vmax)\u001b[0m\n\u001b[1;32m     24\u001b[0m     \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     25\u001b[0m         \u001b[0mcol\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0.3\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m     \u001b[0max\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mY\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmarker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'o'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcol\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mzorder\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     27\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     28\u001b[0m def visualize_network(weights,biases,activations,\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/__init__.py\u001b[0m in \u001b[0;36minner\u001b[0;34m(ax, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1708\u001b[0m                     warnings.warn(msg % (label_namer, func.__name__),\n\u001b[1;32m   1709\u001b[0m                                   RuntimeWarning, stacklevel=2)\n\u001b[0;32m-> 1710\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1711\u001b[0m         \u001b[0mpre_doc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0minner\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__doc__\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1712\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mpre_doc\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/axes/_axes.py\u001b[0m in \u001b[0;36mscatter\u001b[0;34m(self, x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, verts, edgecolors, **kwargs)\u001b[0m\n\u001b[1;32m   4124\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4125\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madd_collection\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcollection\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4126\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautoscale_view\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4127\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4128\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mcollection\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36mautoscale_view\u001b[0;34m(self, tight, scalex, scaley)\u001b[0m\n\u001b[1;32m   2311\u001b[0m         handle_single_axis(\n\u001b[1;32m   2312\u001b[0m             \u001b[0mscaley\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_autoscaleYon\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_shared_y_axes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'intervaly'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2313\u001b[0;31m             'minposy', self.yaxis, self._ymargin, y_stickies, self.set_ybound)\n\u001b[0m\u001b[1;32m   2314\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2315\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_get_axis_list\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36mhandle_single_axis\u001b[0;34m(scale, autoscaleon, shared_axes, interval, minpos, axis, margin, stickies, set_bound)\u001b[0m\n\u001b[1;32m   2302\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2303\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_tight\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2304\u001b[0;31m                 \u001b[0mx0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlocator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mview_limits\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2305\u001b[0m             \u001b[0mset_bound\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2306\u001b[0m             \u001b[0;31m# End of definition of internal function 'handle_single_axis'.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/ticker.py\u001b[0m in \u001b[0;36mview_limits\u001b[0;34m(self, dmin, dmax)\u001b[0m\n\u001b[1;32m   1980\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1981\u001b[0m         dmin, dmax = mtransforms.nonsingular(\n\u001b[0;32m-> 1982\u001b[0;31m             dmin, dmax, expander=1e-12, tiny=1e-13)\n\u001b[0m\u001b[1;32m   1983\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1984\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mrcParams\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'axes.autolimit_mode'\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'round_numbers'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/matplotlib/transforms.py\u001b[0m in \u001b[0;36mnonsingular\u001b[0;34m(vmin, vmax, expander, tiny, increasing)\u001b[0m\n\u001b[1;32m   2892\u001b[0m     \u001b[0mclose\u001b[0m \u001b[0mto\u001b[0m \u001b[0mzero\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mit\u001b[0m \u001b[0mreturns\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mexpander\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mexpander\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2893\u001b[0m     '''\n\u001b[0;32m-> 2894\u001b[0;31m     \u001b[0;32mif\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;32mnot\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misfinite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvmin\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;32mnot\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misfinite\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mvmax\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2895\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0;34m-\u001b[0m\u001b[0mexpander\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexpander\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2896\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "def my_target(y):\n",
    "    a=0.8; r=0.5; R=2.0\n",
    "    return( 1.0*( y[:,0]**2+y[:,1]**2<R**2 ) - 1.0*( (y[:,0]-a)**2+(y[:,1]-a)**2<r**2) - 1.0*( (y[:,0]+a)**2+(y[:,1]-a)**2<r**2 ) )\n",
    "\n",
    "\n",
    "visualize_network_training(weights=[],biases=[],num_neurons=[2,30,30,1],\n",
    "    target_function=my_target, # the target function to approximate\n",
    "    activations=[ 'reLU','reLU','linear' ],\n",
    "    y0range=[-3,3],y1range=[-3,3],\n",
    "    steps=5000, eta=.1, batchsize=200,\n",
    "                          visualize_nsteps=100, \n",
    "                           plot_target=True,\n",
    "                          optimizer='adam',\n",
    "                          size=20,linewidth=2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exercise: Extend this code so as to be able to use more advanced keras activation functions for the layers!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}