{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Important: This notebook will only work with fastai-0.7.x. Do not try to run any fastai-1.x code from this path in the repository because it will load fastai-0.7.x**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- leaky relu / elu"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Planet Kaggle competition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fast_gen import *\n",
    "from learner import *\n",
    "from pt_models import *\n",
    "from dataset_pt import *\n",
    "from sgdr_pt import *\n",
    "\n",
    "from planet import *\n",
    "\n",
    "bs=64; f_model = resnet34\n",
    "path = \"/data/jhoward/fast/planet/\"\n",
    "torch.cuda.set_device(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "n=len(list(open(f'{path}train_v2.csv')))-1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=get_data_pad(f_model, path, 256, 64, n, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = Learner.pretrained_convnet(f_model, data, metrics=[f2])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3c098a7d73ee411f897189755f9952b1"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6c63b18670134e259f70d069524c896d"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.15438985784802958, 0.13450478, 0.87891029153421008]\n"
     ]
    }
   ],
   "source": [
    "learn.fit(0.2, 1, cycle_len=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAD8CAYAAAB3u9PLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f14b71c7fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c1b1d06a508846ad9b5f0fec98d22148"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9333c7ad592e43b7b4f8522e97717380"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.11952567681830356, 0.10556465, 0.90650807074212403]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "abf76c56cba547298155dcf7fc1d7c8c"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.11009075188366914, 0.096736901, 0.91672481730802124]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "771e68e41dea42caa710b507efe0f042"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.10827947736338327, 0.093592748, 0.91964642671839225]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4fb6a0b704da4c048cc6d6bb714a3e9f"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.1058942063357304, 0.092603102, 0.92130033558366631]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ecfb3d14966d4efd96ee857214c80653"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.1092922736445057, 0.095529161, 0.91819015894496181]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a16e83ccd0e0411e96e61ac6a6006fa4"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.1011713251614138, 0.092735209, 0.92159361463751932]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51014344d6024694b6f64ee2f147b0de"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09583996319032348, 0.09039887, 0.92251825740384097]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aa095645f9414808b5df02f83bab7aa7"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09977698443701977, 0.089664221, 0.92254781482418324]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a949819032244b9b44e8162eeead507"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.10362990970969754, 0.092073925, 0.92141614562971053]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5fe6da27198642eda20b8917c06ae61f"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.10036470544572228, 0.090630345, 0.92154762597476325]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6720c350ee01422f840b01e4a353fd43"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09469915216101382, 0.089381441, 0.92234579315867105]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0d997ff6f9f941f98909a7acfd0249f2"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0912562884700505, 0.089002736, 0.92252729549243617]\n"
     ]
    }
   ],
   "source": [
    "learn.fit([0.01,0.05,0.2], 12, cycle_len=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e9e9f06f2e114221990ae001c45376ab"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "97f51d39f7dc40b0918ec2cc5f2f6df5"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09916652571222209, 0.090306021, 0.9227976239444905]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7933fd2e89344e158c2da8644e59575f"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09459107546738892, 0.089821696, 0.92330719115188997]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "88041e8d48c54629b1485c05e48e6e44"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0965355865038755, 0.089525335, 0.92353226343888395]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a9a1e3644dde490dbe41b38df40f2df0"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.09088765592814434, 0.089346372, 0.92387145358147427]\n"
     ]
    }
   ],
   "source": [
    "learn.fit([1e-4,1e-3,0.01], 4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "name = '170809'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_cycle_cv(cv, cycle):\n",
    "    data=get_data_zoom(f_model, path, 256, 64, n, cv)\n",
    "    learn.set_data(data)\n",
    "    learn.load_cycle(f'{name}_{cv}', cycle)\n",
    "    return data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = load_cycle_cv(0,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "val = learn.predict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.92719935814702104"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f2(val,data.val_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.92776745163594243"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f2(learn.TTA(),data.val_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93020923622107132"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f2(val,data.val_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93146299154229428"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f2(learn.TTA(),data.val_y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['partly_cloudy', 'primary']\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQAAAAEACAIAAADTED8xAADY2klEQVR4nMz964IjuY00gEaAqvae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\n",
      "text/plain": [
       "<PIL.Image.Image image mode=RGB size=256x256 at 0x7F284BF8FE80>"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_labels(a): return [data.classes[o] for o in a.nonzero()[0]]\n",
    "\n",
    "lbls = test>0.2\n",
    "idx=9\n",
    "print(get_labels(lbls[idx]))\n",
    "PIL.Image.open(path+data.test_dl.dataset.fnames[idx]).convert('RGB')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['test-jpg/test_17555.jpg',\n",
       " 'test-jpg/test_27899.jpg',\n",
       " 'test-jpg/test_21010.jpg',\n",
       " 'test-jpg/test_39610.jpg',\n",
       " 'test-jpg/test_34768.jpg']"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = [get_labels(o) for o in lbls]\n",
    "data.test_dl.dataset.fnames[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>image_name</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>test_17555</td>\n",
       "      <td>agriculture haze primary road water</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>test_27899</td>\n",
       "      <td>agriculture clear partly_cloudy primary road w...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>test_21010</td>\n",
       "      <td>agriculture bare_ground clear primary road</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>test_39610</td>\n",
       "      <td>agriculture clear primary road</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>test_34768</td>\n",
       "      <td>agriculture clear primary road</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   image_name                                               tags\n",
       "0  test_17555                agriculture haze primary road water\n",
       "1  test_27899  agriculture clear partly_cloudy primary road w...\n",
       "2  test_21010         agriculture bare_ground clear primary road\n",
       "3  test_39610                     agriculture clear primary road\n",
       "4  test_34768                     agriculture clear primary road"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "outp = pd.DataFrame({'image_name': [f[9:-4] for f in data.test_dl.dataset.fnames],\n",
    "        'tags': [' '.join(l) for l in res]})\n",
    "outp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "outp.to_csv('tmp/subm.gz', compression='gzip', index=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import FileLink"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<a href='tmp/subm.gz' target='_blank'>tmp/subm.gz</a><br>"
      ],
      "text/plain": [
       "/data/jhoward/github/deeplearning/nbs/tmp/subm.gz"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FileLink('tmp/subm.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cycle_preds(name, cycle, n_tta=4, is_test=False):\n",
    "    learn.load_cycle(name, cycle)\n",
    "    return learn.TTA(n_tta, is_test=is_test)\n",
    "\n",
    "def cycle_cv_preds(cv, n_tta=4, is_test=False):\n",
    "    data=get_data_pad(f_model, path, 256, 64, n, cv)\n",
    "    learn.set_data(data)\n",
    "    return [cycle_preds(f'{name}_{cv}',i, is_test=is_test) for i in range(5)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- check dogs and cats\n",
    "- get resize working again with new path structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "CPU times: user 6h 23min 45s, sys: 16min 59s, total: 6h 40min 44s\n",
      "Wall time: 6h 43min 38s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "preds_arr=[]\n",
    "for i in range(5):\n",
    "    print(i)\n",
    "    preds_arr.append(cycle_cv_preds(i, is_test=True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def all_cycle_cv_preds(end_cycle, start_cycle=0, n_tta=4, is_test=False):\n",
    "    return [cycle_cv_preds(i, is_test=is_test) for i in range(start_cycle, end_cycle)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.savez_compressed(f'{path}tmp/test_preds', preds_arr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds_avg = [np.mean(o,0) for o in preds_arr]\n",
    "test = np.mean(preds_avg,0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 54min 11s, sys: 2min 40s, total: 56min 52s\n",
      "Wall time: 56min 18s\n"
     ]
    }
   ],
   "source": [
    "%time preds_arr = all_cycle_cv_preds(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.93071536605914396,\n",
       " 0.93116843932677196,\n",
       " 0.93069579784787815,\n",
       " 0.9307239826054069,\n",
       " 0.93051232753450475]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[f2(preds_arr[0][o],data.val_y) for o in range(5)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "preds_avg = [np.mean(o,0) for o in preds_arr]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ys = [get_data_zoom(f_model, path, 256, 64, n, cv).val_y for cv in range(5)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.93137364725177285,\n",
       " 0.93053278668158479,\n",
       " 0.93297669347979839,\n",
       " 0.9323465548921448,\n",
       " 0.93621843040962094]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f2s = [f2(o,y) for o,y in zip(preds_avg,ys)]; f2s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.20000000000000004,\n",
       " 0.17000000000000001,\n",
       " 0.21000000000000005,\n",
       " 0.23000000000000007,\n",
       " 0.19000000000000003]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ots = [opt_th(o,y) for o,y in zip(preds_avg,ys)]; ots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.20000000000000004"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(ots)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93268962254298438"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(f2s,0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### End"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}