{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import unittest\n",
    "from config import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.vision import *\n",
    "from fastai.gen_doc.doctest import this_tests\n",
    "from fastai.callbacks import *\n",
    "from math import isclose\n",
    "from fastai.train import ClassificationInterpretation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = untar_data(URLs.MNIST_TINY)\n",
    "data = ImageDataBunch.from_folder(path, ds_tfms=(rand_pad(2, 28), []), num_workers=2)\n",
    "data = data.normalize()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = Learner(data, simple_cnn((3,16,16,16,2), bn=True), metrics=[accuracy, error_rate])\n",
    "learn.fit_one_cycle(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class IntegrationTest(unittest.TestCase):\n",
    "\n",
    "    def test_accuracy(self):\n",
    "        self.assertGreater(accuracy(*learn.get_preds()), 0.9)\n",
    "        \n",
    "    def test_error_rate(self):\n",
    "        self.assertLess(error_rate(*learn.get_preds()), 0.1)\n",
    "        \n",
    "    def test_preds(self):\n",
    "        pass_tst = False\n",
    "        for i in range(3):\n",
    "            img, label = learn.data.valid_ds[i]\n",
    "            pred_class,pred_idx,outputs = learn.predict(img)\n",
    "            if outputs[int(label)] > outputs[1-int(label)]:\n",
    "                pass_tst = True\n",
    "                break\n",
    "        self.assertTrue(pass_tst)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class OneCycleTest(unittest.TestCase):\n",
    "    \n",
    "    def test_lrs(self):\n",
    "        lrs = learn.recorder.lrs\n",
    "        self.assertLess(lrs[0], 0.001)\n",
    "        self.assertLess(lrs[-1], 0.0001)\n",
    "        self.assertEqual(np.max(lrs), 3e-3)\n",
    "        \n",
    "    def test_moms(self):\n",
    "        moms = learn.recorder.moms\n",
    "        self.assertEqual(moms[0],0.95)\n",
    "        self.assertLess(abs(moms[-1]-0.95), 0.01)\n",
    "        self.assertEqual(np.min(moms),0.85)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}