secondary_eval.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\GitProjects\\Bio-comp\\Coursework\\ANNModel\\model.py:5: TqdmExperimentalWarning: Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)\n",
      "  from tqdm.autonotebook import tqdm\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from Coursework.ANNModel import model as mod\n",
    "from Coursework.ANNModel import data as d\n",
    "from Coursework.PSO import pswarm as ps\n",
    "from Coursework.PSO import psobehaviour as psh\n",
    "from Coursework.PSO import history as hs\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cubic (10 runs) mean fitness: 44.37869624714425\n",
    "Cubic_pso_config_mean = [91.38785822,  6.52696304,  0.59416449,  0.43559926,  0.29415449,\n",
    "        1.29923762,  0.73352514]\n",
    "\n",
    "# Linear (10 runs) mean fintess: 1103190071387.8752\n",
    "Linear_pso_config_mean = [85.37707185,  6.86909089,  0.51382895,  0.9798064,  0.12477974, 0.59493009,  0.89702589]\n",
    "\n",
    "# Tanh (10 runs) mean fitness: 25.520455427936067\n",
    "Tanh_pso_config_mean = [55.32621771,  7.52578441,  0.12798331,  0.72531664,  1.08313114, 0.29216353,  1.3627685 ]\n",
    "\n",
    "# Sine (10 runs) mean fitness: 434\n",
    "Sine_pso_config_mean = [8.73729926e+01, 7.54895398e+00, 7.33998847e-01, 7.52228076e-02,\n",
    "       2.92737853e-01, 3.09977124e-01, 8.51531051e-01]\n",
    "\n",
    "# Complex (10 runs) mean fitness: 12\n",
    "Complex_pso_config_mean = [82.66312781,  5.85540776,  0.62559032,  0.58225519,  0.26541185,\n",
    "        1.38181036,  0.6510299 ]\n",
    "\n",
    "# Xor (10 runs) mean fitness: 2710748526\n",
    "Xor_pso_config_mean = [96.15250011,  7.08445513,  0.53364588,  0.62468836,  1.57375821,\n",
    "        1.50229261,  0.17895104]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cubic Most fit model: 109.50653486975476\n",
    "Cubic_ann_params = np.array([ 5.3012172 , -0.57272851, -0.29479981,  0.80752501, -0.45059291,\n",
    "       -0.37421024, -0.75524689, -0.44855557,  0.89593846,  0.21242254,\n",
    "        0.65222056, -0.53284427, -0.79926862, -0.79949714,  0.96237351])\n",
    "\n",
    "# Linear Most fit model: 10965948279221.076\n",
    "Linear_ann_params = np.array([ 0.15520314,  0.2222954 ,  0.47238101, -0.83439247, -0.27726962,\n",
    "        0.11339216, -0.7816949 , -0.79196132, -0.3392678 ,  0.41548984,\n",
    "       -0.08055252,  0.12817675, -0.71425698, -0.21576561, -0.7553213 ])\n",
    "\n",
    "# Tanh Most fit model: 39.38730853575883\n",
    "Tanh_ann_params = np.array([ 0.09854553,  0.39395654, -0.33737974, -0.12358008,  0.4334227 ,\n",
    "       -0.98235414, -0.99205521,  0.98638961, -0.97515305,  2.33845816,\n",
    "        0.51783499, -0.96708832, -0.9922467 ,  0.9696207 , -0.99365926])\n",
    "\n",
    "# Sine Most fit model: 4212\n",
    "Sine_ann_params = np.array([ 0.26515734,  0.7400037 , -0.0499965 ,  0.32639921,  0.03951493,\n",
    "        0.93575309,  0.82658881,  0.95566027,  0.89756768,  3.01482873,\n",
    "       -0.62419448, -0.9322514 , -0.99330274, -0.85322129, -0.65752741])\n",
    "\n",
    "# Complex Most fit model: 15.36\n",
    "Complex_ann_params = np.array([ 5.16794124, -0.50701965, -0.02085653, -0.67915966, -0.56917589,\n",
    "        0.93068562,  0.25302176,  0.92653444,  0.67209426,  0.39299595,\n",
    "       -0.75832145, -0.44806736, -0.62409455,  2.26094952, -0.26506833,\n",
    "       -0.86461154,  0.31096583,  0.90781783,  0.72454439])\n",
    "\n",
    "# Xor Most fit model: 27107485189\n",
    "Xor_ann_params = np.array([-0.13057076, -0.66250843,  0.0122888 , -0.4294471 , -0.56088656,\n",
    "        0.76634383,  0.95302873,  0.27348615,  0.86272884,  0.79509781,\n",
    "        0.80510636,  0.08453014,  0.86190117,  3.37130705, -0.79043449,\n",
    "        0.74686407,  0.4886179 , -0.22842822,  0.69471657])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = d.Data(\"./Data/1in_cubic.txt\")\n",
    "decimals = 'cubic'\n",
    "X = data.get_rows()\n",
    "y = data.get_output()\n",
    "\n",
    "psoParams = Cubic_pso_config_mean\n",
    "annParams = Cubic_ann_params\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=14)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Compiled!\n",
      "Model Compiled!\n",
      "Model Compiled!\n",
      "Model Compiled!\n"
     ]
    }
   ],
   "source": [
    "import copy\n",
    "base_model = mod.ANN()\n",
    "base_model.set_training_input(X_train, y_train)\n",
    "base_model.add(mod.Layer(1, activation=\"sigmoid\"))\n",
    "base_model.add(mod.Layer(50, activation=\"sigmoid\"))\n",
    "base_model.add(mod.Layer(50, activation=\"sigmoid\"))\n",
    "base_model.add(mod.Layer(1, activation=\"null\"))\n",
    "base_model.compile()\n",
    "\n",
    "#train the model\n",
    "train_model = copy.deepcopy(base_model)\n",
    "train_model.compile()\n",
    "\n",
    "test_model = copy.deepcopy(base_model)\n",
    "test_model.compile()\n",
    "\n",
    "local_best_model = copy.deepcopy(base_model)\n",
    "local_best_model.compile()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sanity check swarm size:  91\n"
     ]
    }
   ],
   "source": [
    "pso = ps.PSO(termination_args={'max_iter': 800}, verbose=True)\n",
    "pso = pso.decode_vec(psoParams)\n",
    "print('Sanity check swarm size: ',pso.swarm_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_model_history = hs.PSOHistory(train_model, num_particles=pso.swarm_size, num_iterations=800)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run experiment with the PSO hyperparameters calculated by meta-PSO \n",
    "#### This will find the 10 most fit ANN parameters, produce a mean accuracy*, Fitness, and loss for both test and train.\n",
    "\n",
    "*This accuracy is calculated with an absolute tolerance of 0.03 for all tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8967064fca5c40399cde3d81fab30575",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "22e1a30ac0f14e5985370a5936361531",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "17886e8fa78347378f69a16c23a4ba05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d9492677652e4c2d86979c99af836084",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "045b5ae2955944178715c15fdf9c8936",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2d5eef8874ec4135b2091163a6906521",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e0be373ff10942eda97cb0fbb90afc91",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cdf85d1f057f4271bfbbae727d65305b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "94fe8b554d9b4ba78cd9a067f2001f07",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2ea5e6b0349b4d75bd0efa23ec7e03c2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(HTML(value='Fitness: -9999.0'), FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Testing on train - A:  0.1044776119402985  B:  0.1044776119402985  Fitness:  34.467  Loss:  0.029\n",
      "Testing on test - A:  0.09411764705882354  B:  0.09411764705882354  Fitness:  44.04  Loss:  0.023\n"
     ]
    }
   ],
   "source": [
    "pso.set_search_dimensions(train_model_history.dimension_vec())\n",
    "pso.set_fitness_fn(train_model_history.evaluate_fitness)\n",
    "\n",
    "train ={\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "test = {\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "for i in range(10):\n",
    "    fitness = pso.run()\n",
    "    # Evaluating on training set\n",
    "    train_model.set_training_input(X_train, y_train)\n",
    "    train['fitnessList'].append(train_model.evaluate_fitness(fitness.location))\n",
    "    score = data.score(train_model, decimals, atol=0.03)\n",
    "    train['scoreA'].append(score[0])\n",
    "    train['scoreB'].append(score[1])\n",
    "    train['lossList'].append(train_model.loss)\n",
    "    # Evaluating on testing set\n",
    "    test_model.set_training_input(X_test, y_test)\n",
    "    test['fitnessList'].append(test_model.evaluate_fitness(fitness.location))\n",
    "    score = data.score(test_model, decimals, atol=0.03)\n",
    "    test['scoreA'].append(score[0])\n",
    "    test['scoreB'].append(score[1])\n",
    "    test['lossList'].append(test_model.loss)\n",
    "\n",
    "print(\"Testing on train - A: \", np.mean(train['scoreA']), \" B: \", np.mean(train['scoreB']), \" Fitness: \", round(np.mean(train['fitnessList']), 3), \" Loss: \", round(np.mean(train['lossList']), 3))\n",
    "print(\"Testing on test - A: \", np.mean(test['scoreA']), \" B: \", np.mean(test['scoreB']), \" Fitness: \", round(np.mean(test['fitnessList']), 3), \" Loss: \", round(np.mean(test['lossList']), 3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Compiled!\n"
     ]
    }
   ],
   "source": [
    "best_model = mod.ANN()\n",
    "best_model.set_training_input(X_train, y_train)\n",
    "best_model.add(mod.Layer(4, activation=\"sigmoid\"))\n",
    "best_model.add(mod.Layer(1, activation=\"null\"))\n",
    "best_model.compile()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing on train - A:  0.13432835820895522  B:  0.13432835820895522  Fitness:  109.507  Loss:  0.009\n",
      "Testing on test - A:  0.17647058823529413  B:  0.17647058823529413  Fitness:  155.279  Loss:  0.006\n"
     ]
    }
   ],
   "source": [
    "best_model.decode_vec(annParams)\n",
    "best_model.one_pass()\n",
    "\n",
    "train ={\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "test = {\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "# Evaluating on training set\n",
    "best_model.set_training_input(X_train, y_train)\n",
    "train['fitnessList'].append(best_model.evaluate_fitness(annParams))\n",
    "score = data.score(best_model, decimals, atol=0.03)\n",
    "train['scoreA'].append(score[0])\n",
    "train['scoreB'].append(score[1])\n",
    "train['lossList'].append(best_model.loss)\n",
    "# Evaluating on testing set\n",
    "best_model.set_training_input(X_test, y_test)\n",
    "test['fitnessList'].append(best_model.evaluate_fitness(annParams))\n",
    "score = data.score(best_model, decimals, atol=0.03)\n",
    "test['scoreA'].append(score[0])\n",
    "test['scoreB'].append(score[1])\n",
    "test['lossList'].append(best_model.loss)\n",
    "\n",
    "print(\"Testing on train - A: \", np.mean(train['scoreA']), \" B: \", np.mean(train['scoreB']), \" Fitness: \", round(np.mean(train['fitnessList']), 3), \" Loss: \", round(np.mean(train['lossList']), 3))\n",
    "print(\"Testing on test - A: \", np.mean(test['scoreA']), \" B: \", np.mean(test['scoreB']), \" Fitness: \", round(np.mean(test['fitnessList']), 3), \" Loss: \", round(np.mean(test['lossList']), 3))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_10_run_best_fitness, mean_10_run_best_location = train_model_history.best_particle()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_particle_fitness_dict = train_model_history.historical_particle_fitness()\n",
    "train_particle_location_dict = train_model_history.historical_particle_location()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_model_history.reshape_lists(10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(None, None)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_model_history.plot_fitness(),train_model_history.plot_mean_fitness()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing on train - A:  0.13432835820895522  B:  0.13432835820895522  Fitness:  39.47  Loss:  0.025\n",
      "Testing on test - A:  0.058823529411764705  B:  0.058823529411764705  Fitness:  52.556  Loss:  0.019\n"
     ]
    }
   ],
   "source": [
    "local_best_model.decode_vec(mean_10_run_best_location)\n",
    "local_best_model.one_pass()\n",
    "\n",
    "train ={\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "test = {\n",
    "    'scoreA' : [],\n",
    "    'scoreB' : [],\n",
    "    'fitnessList': [],\n",
    "    'lossList': []\n",
    "}\n",
    "\n",
    "# Evaluating on training set\n",
    "local_best_model.set_training_input(X_train, y_train)\n",
    "train['fitnessList'].append(local_best_model.evaluate_fitness(mean_10_run_best_location))\n",
    "score = data.score(local_best_model, decimals, atol=0.03)\n",
    "train['scoreA'].append(score[0])\n",
    "train['scoreB'].append(score[1])\n",
    "train['lossList'].append(local_best_model.loss)\n",
    "# Evaluating on testing set\n",
    "local_best_model.set_training_input(X_test, y_test)\n",
    "test['fitnessList'].append(local_best_model.evaluate_fitness(mean_10_run_best_location))\n",
    "score = data.score(local_best_model, decimals, atol=0.03)\n",
    "test['scoreA'].append(score[0])\n",
    "test['scoreB'].append(score[1])\n",
    "test['lossList'].append(local_best_model.loss)\n",
    "\n",
    "print(\"Testing on train - A: \", np.mean(train['scoreA']), \" B: \", np.mean(train['scoreB']), \" Fitness: \", round(np.mean(train['fitnessList']), 3), \" Loss: \", round(np.mean(train['lossList']), 3))\n",
    "print(\"Testing on test - A: \", np.mean(test['scoreA']), \" B: \", np.mean(test['scoreB']), \" Fitness: \", round(np.mean(test['fitnessList']), 3), \" Loss: \", round(np.mean(test['lossList']), 3))\n"
   ]
  },
  {
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}