PointNet2_xyz_sd.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "be8RFvYTsfYE",
    "tags": []
   },
   "source": [
    "# Data Load\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "executionInfo": {
     "elapsed": 15376,
     "status": "ok",
     "timestamp": 1674079742236,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "0J4zJlrCqSb4"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import math\n",
    "import random\n",
    "import os\n",
    "import torch\n",
    "import scipy.spatial.distance\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torchvision import transforms, utils\n",
    "\n",
    "import plotly.graph_objects as go\n",
    "import plotly.express as px\n",
    "\n",
    "import matplotlib.pylab as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import pandas as pd\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "# from google.colab import drive\n",
    "import tensorflow as tf\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 8394,
     "status": "ok",
     "timestamp": 1677260843122,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "wOHAa2Ku_ci8",
    "outputId": "1c64c450-9b00-4473-c57b-e6e7396dccd0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1     2869200\n",
      "2      826645\n",
      "5      746775\n",
      "12     378061\n",
      "16     366536\n",
      "17     209118\n",
      "3      111354\n",
      "15      35457\n",
      "11      22421\n",
      "9        7342\n",
      "13       1932\n",
      "6        1319\n",
      "14        731\n",
      "Name: cla, dtype: int64\n",
      "1     417280\n",
      "2     276623\n",
      "16    177257\n",
      "5      93060\n",
      "13     41630\n",
      "17     24679\n",
      "12     14055\n",
      "3       4101\n",
      "4       1072\n",
      "9        682\n",
      "Name: cla, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "lxyz = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959322/DEBY_LOD2_4959322.xyz', sep=\" \", header=0)\n",
    "lxyz.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "print(lxyz.cla.value_counts())\n",
    "\n",
    "lxyz_test = pd.read_csv('./data/test/labelledLocalCRS/DEBY_LOD2_4959459/DEBY_LOD2_4959459.xyz', sep=\" \", header=0)\n",
    "lxyz_test.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "print(lxyz_test.cla.value_counts())\n",
    "\n",
    "############## !! Here you can choose whether to use large dataset ###############"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 10506,
     "status": "ok",
     "timestamp": 1674079752704,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "wOHAa2Ku_ci8",
    "outputId": "aa11cb6d-3ffe-434d-da6c-309d1d541783"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1     36429131\n",
      "16     7566216\n",
      "11     7534665\n",
      "5      7056168\n",
      "2      6351914\n",
      "12     5938935\n",
      "15     3797046\n",
      "17     3600867\n",
      "7      1393392\n",
      "3       691885\n",
      "6       475590\n",
      "10      419409\n",
      "14      304608\n",
      "8       220774\n",
      "13       49393\n",
      "9        29398\n",
      "Name: cla, dtype: int64\n",
      "1     19125652\n",
      "5      6440977\n",
      "2      3448050\n",
      "12     1979855\n",
      "16     1911652\n",
      "17     1746219\n",
      "6       628964\n",
      "3       288073\n",
      "14      242680\n",
      "13       24642\n",
      "Name: cla, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# lxyz = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959322/DEBY_LOD2_4959322.xyz', sep=\" \", header=0)\n",
    "# lxyz.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "lxyz_1 = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4906981/DEBY_LOD2_4906981.xyz', sep=\" \", header=0)\n",
    "lxyz_1.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "lxyz_2 = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959322/DEBY_LOD2_4959322.xyz', sep=\" \", header=0)\n",
    "lxyz_2.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "lxyz_3 = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959460/DEBY_LOD2_4959460.xyz', sep=\" \", header=0)\n",
    "lxyz_3.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "lxyz_4 = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959462/DEBY_LOD2_4959462.xyz', sep=\" \", header=0)\n",
    "lxyz_4.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "lxyz = pd.concat([lxyz_1, lxyz_2, lxyz_3, lxyz_4], axis = 0)\n",
    "print(lxyz.cla.value_counts())\n",
    "\n",
    "lxyz_test = pd.read_csv('./data/training/labelledLocalCRS/DEBY_LOD2_4959323/DEBY_LOD2_4959323.xyz', sep=\" \", header=0)\n",
    "lxyz_test.columns = [\"x\", \"y\", \"z\", \"cla\"]\n",
    "print(lxyz_test.cla.value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Sample Points & Merge class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1     2911999\n",
      "2      827376\n",
      "5      748094\n",
      "13     577586\n",
      "11     400482\n",
      "3      111354\n",
      "Name: cla, dtype: int64\n",
      "1     419034\n",
      "2     276623\n",
      "13    243566\n",
      "5      93060\n",
      "11     14055\n",
      "3       4101\n",
      "Name: cla, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "lclass = lxyz['cla'].replace({6:5,9:1,15:1,10:1,12:11,17:13,16:13,14:2,4:1})\n",
    "print(lclass.value_counts())\n",
    "\n",
    "lclass_test = lxyz_test['cla'].replace({6:5,9:1,15:1,10:1,12:11,17:13,16:13,14:2,4:1})\n",
    "print(lclass_test.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1    2911999\n",
      "2     827376\n",
      "4     748094\n",
      "5     577586\n",
      "6     400482\n",
      "3     111354\n",
      "Name: cla, dtype: int64\n",
      "1    419034\n",
      "2    276623\n",
      "5    243566\n",
      "4     93060\n",
      "6     14055\n",
      "3      4101\n",
      "Name: cla, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "lclass = lclass.replace({5:4,13:5,11:6})\n",
    "print(lclass.value_counts())\n",
    "# 1: wall\n",
    "# 2: window\n",
    "# 3: door\n",
    "# 4: molding\n",
    "# 5: other\n",
    "# 6: terrian\n",
    "# 7: column\n",
    "# 8: arch\n",
    "\n",
    "lclass_test = lclass_test.replace({5:4,13:5,11:6})\n",
    "print(lclass_test.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    2911999\n",
      "1     827376\n",
      "3     748094\n",
      "4     577586\n",
      "5     400482\n",
      "2     111354\n",
      "Name: cla, dtype: int64\n",
      "0    419034\n",
      "1    276623\n",
      "4    243566\n",
      "3     93060\n",
      "5     14055\n",
      "2      4101\n",
      "Name: cla, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "lclass = lclass.replace({1:0,2:1,3:2,4:3,5:4,6:5,7:6,8:7})\n",
    "print(lclass.value_counts())\n",
    "# 0: wall\n",
    "# 1: window\n",
    "# 2: door\n",
    "# 3: molding\n",
    "# 4: other\n",
    "# 5: terrian\n",
    "# 6: column\n",
    "# 7: arch\n",
    "\n",
    "lclass_test = lclass_test.replace({1:0,2:1,3:2,4:3,5:4,6:5,7:6,8:7})\n",
    "print(lclass_test.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "lxyz = lxyz.drop('cla',axis = 1) \n",
    "lxyz_test = lxyz_test.drop('cla',axis = 1) \n",
    "\n",
    "lxyz = lxyz.to_numpy().reshape(-1,3)\n",
    "lxyz_test = lxyz_test.to_numpy().reshape(-1,3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "import open3d as o3d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "######################### using open3d for point processing ################\n",
    "device = o3d.core.Device(\"CPU:0\")\n",
    "dtype = o3d.core.float32\n",
    "pcd = o3d.t.geometry.PointCloud(device)\n",
    "pcd.point.positions = o3d.core.Tensor(lxyz, dtype, device)\n",
    "pcd.point.labels = o3d.core.Tensor(lclass.to_numpy().reshape(-1,1), o3d.core.int32, device)\n",
    "# o3d.visualization.draw([pcd])\n",
    "downpcd = pcd.voxel_down_sample(voxel_size = 0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "lxyz_train = downpcd.point.positions.numpy()\n",
    "lclass_train = downpcd.point.labels.numpy()\n",
    "\n",
    "pcd_train = o3d.geometry.PointCloud()\n",
    "pcd_train.points = o3d.utility.Vector3dVector(lxyz_train)\n",
    "\n",
    "# color\n",
    "colors = plt.get_cmap(\"tab20\")(np.array(lclass_train).reshape(-1) / 17.0)\n",
    "colors = colors[:,0:3]\n",
    "pcd_train.colors=o3d.utility.Vector3dVector(colors)\n",
    "o3d.visualization.draw_geometries([pcd_train], window_name='test the color',width=800,height=600)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1241136\n"
     ]
    }
   ],
   "source": [
    "print(len(lxyz_train)) # see how many points after simplified"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "pcd_test = o3d.t.geometry.PointCloud(device)\n",
    "pcd_test.point.positions = o3d.core.Tensor(lxyz_test, dtype, device)\n",
    "pcd_test.point.labels = o3d.core.Tensor(lclass_test.to_numpy().reshape(-1,1), o3d.core.int32, device)\n",
    "# o3d.visualization.draw([pcd])\n",
    "downpcd_test = pcd_test.voxel_down_sample(voxel_size = 0.05)\n",
    "\n",
    "lxyz_test = downpcd_test.point.positions.numpy()\n",
    "lclass_test = downpcd_test.point.labels.numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "pcd_test = o3d.geometry.PointCloud()\n",
    "pcd_test.points = o3d.utility.Vector3dVector(lxyz_test)\n",
    "\n",
    "# color\n",
    "colors = plt.get_cmap(\"tab20\")(np.array(lclass_test).reshape(-1) / 17.0)\n",
    "colors = colors[:,0:3]\n",
    "pcd_test.colors=o3d.utility.Vector3dVector(colors)\n",
    "o3d.visualization.draw_geometries([pcd_test], window_name='test the color',width=800,height=600)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Add Features （ IF NOT KEEP AS X Y Z）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "lxyz_train = pd.DataFrame(lxyz_train)\n",
    "lxyz_train.columns = [\"x\", \"y\", \"z\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "lxyz_test = pd.DataFrame(lxyz_test)\n",
    "lxyz_test.columns = [\"x\", \"y\", \"z\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "########## Normalization For Each Column ##########\n",
    "def normalize_point_cloud(pcd):\n",
    "    centroid = np.mean(pcd, axis=0) # calculation the center of point cloud\n",
    "    pcd = pcd - centroid # put the center as origin\n",
    "    m = np.max(np.sqrt(np.sum(pcd ** 2, axis=1))) # calculation the longest vector\n",
    "    pcd_normalized = pcd / m # normalization to (-1, 1)\n",
    "    return pcd_normalized\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "norm_lxyz_train = normalize_point_cloud(lxyz_train)\n",
    "norm_lxyz_test = normalize_point_cloud(lxyz_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "lclass_train = pd.DataFrame(lclass_train, columns = ['cla'], dtype = int)\n",
    "lclass_test = pd.DataFrame(lclass_test, columns = ['cla'], dtype = int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cla\n",
      "0      232389\n",
      "4      173752\n",
      "1      149798\n",
      "3       56679\n",
      "5       12296\n",
      "2        2676\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(lclass_test.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "         cla\n",
      "0          0\n",
      "1          0\n",
      "2          0\n",
      "3          0\n",
      "4          0\n",
      "...      ...\n",
      "1241131    4\n",
      "1241132    4\n",
      "1241133    4\n",
      "1241134    4\n",
      "1241135    4\n",
      "\n",
      "[1241136 rows x 1 columns]\n"
     ]
    }
   ],
   "source": [
    "print(lclass_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "        cla\n",
      "0         4\n",
      "1         4\n",
      "2         4\n",
      "3         4\n",
      "4         4\n",
      "...     ...\n",
      "627585    4\n",
      "627586    4\n",
      "627587    4\n",
      "627588    4\n",
      "627589    4\n",
      "\n",
      "[627590 rows x 1 columns]\n"
     ]
    }
   ],
   "source": [
    "print(lclass_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                x         y         z\n",
      "0        0.055158  0.046167  0.145007\n",
      "1        0.055158  0.048114  0.098288\n",
      "2        0.055158  0.059794 -0.195654\n",
      "3        0.055158  0.061740 -0.219013\n",
      "4        0.055158  0.053954 -0.170347\n",
      "...           ...       ...       ...\n",
      "1241131 -0.055800  0.098726 -0.000990\n",
      "1241132 -0.055800  0.100673  0.012636\n",
      "1241133 -0.055800  0.096780 -0.000990\n",
      "1241134 -0.472380  0.273923 -0.006830\n",
      "1241135 -0.480166  0.268084 -0.076909\n",
      "\n",
      "[1241136 rows x 3 columns]\n"
     ]
    }
   ],
   "source": [
    "print(norm_lxyz_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "               x         y         z\n",
      "0       0.476475  0.267497 -0.053678\n",
      "1       0.475024  0.264596  0.294546\n",
      "2       0.475024  0.263145  0.294546\n",
      "3       0.475024  0.267497  0.295997\n",
      "4       0.475024  0.258792  0.307604\n",
      "...          ...       ...       ...\n",
      "627585 -0.337498 -0.459420 -0.214731\n",
      "627586 -0.337498 -0.460871 -0.214731\n",
      "627587 -0.337498 -0.455067 -0.216182\n",
      "627588 -0.338949 -0.457969 -0.214731\n",
      "627589 -0.338949 -0.449263 -0.214731\n",
      "\n",
      "[627590 rows x 3 columns]\n"
     ]
    }
   ],
   "source": [
    "print(norm_lxyz_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Data Spilt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "lxyz_train = pd.concat([norm_lxyz_train, lclass_train], axis = 1)\n",
    "lxyz_test = pd.concat([norm_lxyz_test, lclass_test], axis = 1)\n",
    "# print(lxyz_test.cla.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                x         y         z  cla\n",
      "0        0.055158  0.046167  0.145007    0\n",
      "1        0.055158  0.048114  0.098288    0\n",
      "2        0.055158  0.059794 -0.195654    0\n",
      "3        0.055158  0.061740 -0.219013    0\n",
      "4        0.055158  0.053954 -0.170347    0\n",
      "...           ...       ...       ...  ...\n",
      "1241131 -0.055800  0.098726 -0.000990    4\n",
      "1241132 -0.055800  0.100673  0.012636    4\n",
      "1241133 -0.055800  0.096780 -0.000990    4\n",
      "1241134 -0.472380  0.273923 -0.006830    4\n",
      "1241135 -0.480166  0.268084 -0.076909    4\n",
      "\n",
      "[1241136 rows x 4 columns]\n",
      "               x         y         z  cla\n",
      "0       0.476475  0.267497 -0.053678    4\n",
      "1       0.475024  0.264596  0.294546    4\n",
      "2       0.475024  0.263145  0.294546    4\n",
      "3       0.475024  0.267497  0.295997    4\n",
      "4       0.475024  0.258792  0.307604    4\n",
      "...          ...       ...       ...  ...\n",
      "627585 -0.337498 -0.459420 -0.214731    4\n",
      "627586 -0.337498 -0.460871 -0.214731    4\n",
      "627587 -0.337498 -0.455067 -0.216182    4\n",
      "627588 -0.338949 -0.457969 -0.214731    4\n",
      "627589 -0.338949 -0.449263 -0.214731    4\n",
      "\n",
      "[627590 rows x 4 columns]\n"
     ]
    }
   ],
   "source": [
    "print(lxyz_train)\n",
    "print(lxyz_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 2139,
     "status": "ok",
     "timestamp": 1674079754823,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "ey4Eh_U10ET5",
    "outputId": "d4aedcb5-0aef-465c-b3ce-f0fd69af7890"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "489\n",
      "132\n",
      "102\n",
      "284\n",
      "188\n",
      "14\n",
      "0    500736\n",
      "4    290816\n",
      "1    192512\n",
      "3    135168\n",
      "5    104448\n",
      "2     14336\n",
      "Name: cla, dtype: int64\n",
      "---------\n",
      "169\n",
      "146\n",
      "226\n",
      "55\n",
      "2\n",
      "12\n"
     ]
    }
   ],
   "source": [
    "def spilt_data(lxyz):\n",
    "  class_uniques = lxyz['cla'].unique() #把一列所有取值提出来成为一个Array \n",
    "  re = []\n",
    "  for column in class_uniques:\n",
    "    temp_data = lxyz[lxyz['cla'].isin([column])] #按照列的某个值取一个dataframe\n",
    "    batch_num = int(len(temp_data)/1024)\n",
    "    print(batch_num)\n",
    "    temp_data = temp_data[0:batch_num*1024]\n",
    "    re.append(temp_data)\n",
    "  re = pd.concat(re)\n",
    "  return re\n",
    "\n",
    "lxyz_train = spilt_data(lxyz_train)\n",
    "print(lxyz_train.cla.value_counts())\n",
    "print('---------')\n",
    "lxyz_test = spilt_data(lxyz_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                x         y         z  cla\n",
      "0        0.055158  0.046167  0.145007    0\n",
      "1        0.055158  0.048114  0.098288    0\n",
      "2        0.055158  0.059794 -0.195654    0\n",
      "3        0.055158  0.061740 -0.219013    0\n",
      "4        0.055158  0.053954 -0.170347    0\n",
      "...           ...       ...       ...  ...\n",
      "1150460  0.331580 -0.076471 -0.121682    2\n",
      "1150461  0.331580 -0.080364 -0.082749    2\n",
      "1150462  0.331580 -0.080364 -0.185921    2\n",
      "1150463  0.331580 -0.076471 -0.131415    2\n",
      "1150464  0.331580 -0.082311 -0.080802    2\n",
      "\n",
      "[1238016 rows x 4 columns]\n",
      "               x         y         z  cla\n",
      "0       0.476475  0.267497 -0.053678    4\n",
      "1       0.475024  0.264596  0.294546    4\n",
      "2       0.475024  0.263145  0.294546    4\n",
      "3       0.475024  0.267497  0.295997    4\n",
      "4       0.475024  0.258792  0.307604    4\n",
      "...          ...       ...       ...  ...\n",
      "612172 -0.334596 -0.452165 -0.214731    5\n",
      "612173 -0.334596 -0.450714 -0.216182    5\n",
      "612174 -0.334596 -0.455067 -0.214731    5\n",
      "612175 -0.334596 -0.439107 -0.214731    5\n",
      "612176 -0.334596 -0.442008 -0.213280    5\n",
      "\n",
      "[624640 rows x 4 columns]\n"
     ]
    }
   ],
   "source": [
    "print(lxyz_train)\n",
    "print(lxyz_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "sSRH6VAGf7Xw",
    "tags": []
   },
   "source": [
    "# Dataset Preparation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "executionInfo": {
     "elapsed": 77,
     "status": "ok",
     "timestamp": 1674079787517,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "n23Bk90QSKqq"
   },
   "outputs": [],
   "source": [
    "class ToTensor(object):\n",
    "    def __call__(self, pointcloud):\n",
    "        assert len(pointcloud.shape)==2\n",
    "        pointcloud = pointcloud.to_numpy()\n",
    "        return torch.from_numpy(pointcloud)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "executionInfo": {
     "elapsed": 77,
     "status": "ok",
     "timestamp": 1674079787518,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "JYsFkTHqfLRp"
   },
   "outputs": [],
   "source": [
    "def default_transforms():\n",
    "    return transforms.Compose([\n",
    "                                # Normalize(),\n",
    "                                ToTensor()\n",
    "                              ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 885,
     "status": "ok",
     "timestamp": 1674079788327,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "Z-Ur3QdPfLX1",
    "outputId": "a7977d44-a242-4069-d2be-eb22720ec8bd"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1238016\n",
      "624640\n",
      "610\n",
      "1209\n",
      "type of :  torch.Size([1024, 4])\n",
      "tensor([[0.4112, 0.3371, 0.3163, 4.0000],\n",
      "        [0.4112, 0.2719, 0.0044, 4.0000],\n",
      "        [0.4112, 0.3255, 0.1030, 4.0000],\n",
      "        ...,\n",
      "        [0.3952, 0.2965, 0.2162, 4.0000],\n",
      "        [0.3952, 0.2965, 0.2264, 4.0000],\n",
      "        [0.3952, 0.2907, 0.2104, 4.0000]], dtype=torch.float64)\n"
     ]
    }
   ],
   "source": [
    "train_transforms = default_transforms()\n",
    "batch_num_test = int(len(lxyz_test)/1024)\n",
    "batch_num_train = int(len(lxyz_train)/1024)\n",
    "_train = lxyz_train[0:(1024*batch_num_train)]\n",
    "_test = lxyz_test[0:(1024*batch_num_test)]\n",
    "print(len(_train))\n",
    "print(len(_test))\n",
    "print(batch_num_test)\n",
    "print(batch_num_train)\n",
    "\n",
    "data_test_temp = np.split(_test,batch_num_test)\n",
    "data_train_temp = np.split(_train,batch_num_train)\n",
    "\n",
    "data_test = [ train_transforms(test) for test in data_test_temp]\n",
    "data_train = [ train_transforms(train) for train in data_train_temp]\n",
    "\n",
    "print('type of : ', data_test[2].size())\n",
    "print(data_test[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1024, 4])\n"
     ]
    }
   ],
   "source": [
    "print(data_train[1208].size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1209\n",
      "906\n",
      "303\n"
     ]
    }
   ],
   "source": [
    "len_total = len(data_train)\n",
    "print(len_total)\n",
    "train_size = int(len_total*0.75)\n",
    "test_size = len_total - train_size\n",
    "print(train_size)\n",
    "print(test_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "303\n",
      "type of :  torch.Size([1024, 4])\n",
      "tensor([[ 0.0357,  0.0715, -0.2171,  0.0000],\n",
      "        [ 0.0357,  0.0540, -0.0983,  0.0000],\n",
      "        [ 0.0357,  0.0656, -0.1256,  0.0000],\n",
      "        ...,\n",
      "        [ 0.0337,  0.0637, -0.0516,  0.0000],\n",
      "        [ 0.0337,  0.0617,  0.0438,  0.0000],\n",
      "        [ 0.0337,  0.0540,  0.1956,  0.0000]], dtype=torch.float64)\n"
     ]
    }
   ],
   "source": [
    "data_train, data_valid = torch.utils.data.random_split(data_train, [train_size, test_size])\n",
    "print(len(data_valid))\n",
    "print('type of : ', data_valid[2].size())\n",
    "print(data_valid[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "executionInfo": {
     "elapsed": 65,
     "status": "ok",
     "timestamp": 1674079788359,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "hNZb5f-K1lrh"
   },
   "outputs": [],
   "source": [
    "train_loader = DataLoader(dataset=data_train, batch_size= 4, shuffle=True)\n",
    "valid_loader = DataLoader(dataset=data_valid, batch_size= 4, shuffle=True)\n",
    "test_loader = DataLoader(dataset=data_test, batch_size = 4)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "N0FRlJOh2EuO",
    "tags": []
   },
   "source": [
    "# Model Description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from time import time\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "def timeit(tag, t):\n",
    "    print(\"{}: {}s\".format(tag, time() - t))\n",
    "    return time()\n",
    "\n",
    "def pc_normalize(pc):\n",
    "    l = pc.shape[0]\n",
    "    centroid = np.mean(pc, axis=0)\n",
    "    pc = pc - centroid\n",
    "    m = np.max(np.sqrt(np.sum(pc**2, axis=1)))\n",
    "    pc = pc / m\n",
    "    return pc\n",
    "\n",
    "def square_distance(src, dst):\n",
    "    \"\"\"\n",
    "    Calculate Euclid distance between each two points.\n",
    "\n",
    "    src^T * dst = xn * xm + yn * ym + zn * zm；\n",
    "    sum(src^2, dim=-1) = xn*xn + yn*yn + zn*zn;\n",
    "    sum(dst^2, dim=-1) = xm*xm + ym*ym + zm*zm;\n",
    "    dist = (xn - xm)^2 + (yn - ym)^2 + (zn - zm)^2\n",
    "         = sum(src**2,dim=-1)+sum(dst**2,dim=-1)-2*src^T*dst\n",
    "\n",
    "    Input:\n",
    "        src: source points, [B, N, C]\n",
    "        dst: target points, [B, M, C]\n",
    "    Output:\n",
    "        dist: per-point square distance, [B, N, M]\n",
    "    \"\"\"\n",
    "    B, N, _ = src.shape\n",
    "    _, M, _ = dst.shape\n",
    "    dist = -2 * torch.matmul(src, dst.permute(0, 2, 1))\n",
    "    dist += torch.sum(src ** 2, -1).view(B, N, 1)\n",
    "    dist += torch.sum(dst ** 2, -1).view(B, 1, M)\n",
    "    return dist\n",
    "\n",
    "\n",
    "def index_points(points, idx):\n",
    "    \"\"\"\n",
    "\n",
    "    Input:\n",
    "        points: input points data, [B, N, C]\n",
    "        idx: sample index data, [B, S]\n",
    "    Return:\n",
    "        new_points:, indexed points data, [B, S, C]\n",
    "    \"\"\"\n",
    "    device = points.device\n",
    "    B = points.shape[0]\n",
    "    view_shape = list(idx.shape)\n",
    "    view_shape[1:] = [1] * (len(view_shape) - 1)\n",
    "    repeat_shape = list(idx.shape)\n",
    "    repeat_shape[0] = 1\n",
    "    batch_indices = torch.arange(B, dtype=torch.long).to(device).view(view_shape).repeat(repeat_shape)\n",
    "    new_points = points[batch_indices, idx, :]\n",
    "    return new_points\n",
    "\n",
    "\n",
    "def farthest_point_sample(xyz, npoint):\n",
    "    \"\"\"\n",
    "    Input:\n",
    "        xyz: pointcloud data, [B, N, 3]\n",
    "        npoint: number of samples\n",
    "    Return:\n",
    "        centroids: sampled pointcloud index, [B, npoint]\n",
    "    \"\"\"\n",
    "    device = xyz.device\n",
    "    B, N, C = xyz.shape\n",
    "    centroids = torch.zeros(B, npoint, dtype=torch.long).to(device)\n",
    "    distance = torch.ones(B, N).to(device) * 1e10\n",
    "    farthest = torch.randint(0, N, (B,), dtype=torch.long).to(device)\n",
    "    batch_indices = torch.arange(B, dtype=torch.long).to(device)\n",
    "    for i in range(npoint):\n",
    "        centroids[:, i] = farthest\n",
    "        centroid = xyz[batch_indices, farthest, :].view(B, 1, 3)\n",
    "        dist = torch.sum((xyz - centroid) ** 2, -1)\n",
    "        mask = dist < distance\n",
    "        distance[mask] = dist[mask]\n",
    "        farthest = torch.max(distance, -1)[1]\n",
    "    return centroids\n",
    "\n",
    "\n",
    "def query_ball_point(radius, nsample, xyz, new_xyz):\n",
    "    \"\"\"\n",
    "    Input:\n",
    "        radius: local region radius\n",
    "        nsample: max sample number in local region\n",
    "        xyz: all points, [B, N, 3]\n",
    "        new_xyz: query points, [B, S, 3]\n",
    "    Return:\n",
    "        group_idx: grouped points index, [B, S, nsample]\n",
    "    \"\"\"\n",
    "    device = xyz.device\n",
    "    B, N, C = xyz.shape\n",
    "    _, S, _ = new_xyz.shape\n",
    "    group_idx = torch.arange(N, dtype=torch.long).to(device).view(1, 1, N).repeat([B, S, 1])\n",
    "    sqrdists = square_distance(new_xyz, xyz)\n",
    "    group_idx[sqrdists > radius ** 2] = N\n",
    "    group_idx = group_idx.sort(dim=-1)[0][:, :, :nsample]\n",
    "    group_first = group_idx[:, :, 0].view(B, S, 1).repeat([1, 1, nsample])\n",
    "    mask = group_idx == N\n",
    "    group_idx[mask] = group_first[mask]\n",
    "    return group_idx\n",
    "\n",
    "\n",
    "def sample_and_group(npoint, radius, nsample, xyz, points, returnfps=False):\n",
    "    \"\"\"\n",
    "    Input:\n",
    "        npoint:\n",
    "        radius:\n",
    "        nsample:\n",
    "        xyz: input points position data, [B, N, 3]\n",
    "        points: input points data, [B, N, D]\n",
    "    Return:\n",
    "        new_xyz: sampled points position data, [B, npoint, nsample, 3]\n",
    "        new_points: sampled points data, [B, npoint, nsample, 3+D]\n",
    "    \"\"\"\n",
    "    B, N, C = xyz.shape\n",
    "    S = npoint\n",
    "    fps_idx = farthest_point_sample(xyz, npoint) # [B, npoint, C]\n",
    "    new_xyz = index_points(xyz, fps_idx)\n",
    "    idx = query_ball_point(radius, nsample, xyz, new_xyz)\n",
    "    grouped_xyz = index_points(xyz, idx) # [B, npoint, nsample, C]\n",
    "    grouped_xyz_norm = grouped_xyz - new_xyz.view(B, S, 1, C)\n",
    "\n",
    "    if points is not None:\n",
    "        grouped_points = index_points(points, idx)\n",
    "        new_points = torch.cat([grouped_xyz_norm, grouped_points], dim=-1) # [B, npoint, nsample, C+D]\n",
    "    else:\n",
    "        new_points = grouped_xyz_norm\n",
    "    if returnfps:\n",
    "        return new_xyz, new_points, grouped_xyz, fps_idx\n",
    "    else:\n",
    "        return new_xyz, new_points\n",
    "\n",
    "\n",
    "def sample_and_group_all(xyz, points):\n",
    "    \"\"\"\n",
    "    Input:\n",
    "        xyz: input points position data, [B, N, 3]\n",
    "        points: input points data, [B, N, D]\n",
    "    Return:\n",
    "        new_xyz: sampled points position data, [B, 1, 3]\n",
    "        new_points: sampled points data, [B, 1, N, 3+D]\n",
    "    \"\"\"\n",
    "    device = xyz.device\n",
    "    B, N, C = xyz.shape\n",
    "    new_xyz = torch.zeros(B, 1, C).to(device)\n",
    "    grouped_xyz = xyz.view(B, 1, N, C)\n",
    "    if points is not None:\n",
    "        new_points = torch.cat([grouped_xyz, points.view(B, 1, N, -1)], dim=-1)\n",
    "    else:\n",
    "        new_points = grouped_xyz\n",
    "    return new_xyz, new_points\n",
    "\n",
    "\n",
    "class PointNetSetAbstraction(nn.Module):\n",
    "    def __init__(self, npoint, radius, nsample, in_channel, mlp, group_all):\n",
    "        super(PointNetSetAbstraction, self).__init__()\n",
    "        self.npoint = npoint\n",
    "        self.radius = radius\n",
    "        self.nsample = nsample\n",
    "        self.mlp_convs = nn.ModuleList()\n",
    "        self.mlp_bns = nn.ModuleList()\n",
    "        last_channel = in_channel\n",
    "        for out_channel in mlp:\n",
    "            self.mlp_convs.append(nn.Conv2d(last_channel, out_channel, 1))\n",
    "            self.mlp_bns.append(nn.BatchNorm2d(out_channel))\n",
    "            last_channel = out_channel\n",
    "        self.group_all = group_all\n",
    "\n",
    "    def forward(self, xyz, points):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            xyz: input points position data, [B, C, N]\n",
    "            points: input points data, [B, D, N]\n",
    "        Return:\n",
    "            new_xyz: sampled points position data, [B, C, S]\n",
    "            new_points_concat: sample points feature data, [B, D', S]\n",
    "        \"\"\"\n",
    "        xyz = xyz.permute(0, 2, 1)\n",
    "        if points is not None:\n",
    "            points = points.permute(0, 2, 1)\n",
    "\n",
    "        if self.group_all:\n",
    "            new_xyz, new_points = sample_and_group_all(xyz, points)\n",
    "        else:\n",
    "            new_xyz, new_points = sample_and_group(self.npoint, self.radius, self.nsample, xyz, points)\n",
    "        # new_xyz: sampled points position data, [B, npoint, C]\n",
    "        # new_points: sampled points data, [B, npoint, nsample, C+D]\n",
    "        new_points = new_points.permute(0, 3, 2, 1) # [B, C+D, nsample,npoint]\n",
    "        for i, conv in enumerate(self.mlp_convs):\n",
    "            bn = self.mlp_bns[i]\n",
    "            new_points =  F.relu(bn(conv(new_points)))\n",
    "\n",
    "        new_points = torch.max(new_points, 2)[0]\n",
    "        new_xyz = new_xyz.permute(0, 2, 1)\n",
    "        return new_xyz, new_points\n",
    "\n",
    "\n",
    "class PointNetSetAbstractionMsg(nn.Module):\n",
    "    def __init__(self, npoint, radius_list, nsample_list, in_channel, mlp_list):\n",
    "        super(PointNetSetAbstractionMsg, self).__init__()\n",
    "        self.npoint = npoint\n",
    "        self.radius_list = radius_list\n",
    "        self.nsample_list = nsample_list\n",
    "        self.conv_blocks = nn.ModuleList()\n",
    "        self.bn_blocks = nn.ModuleList()\n",
    "        for i in range(len(mlp_list)):\n",
    "            convs = nn.ModuleList()\n",
    "            bns = nn.ModuleList()\n",
    "            last_channel = in_channel + 3\n",
    "            for out_channel in mlp_list[i]:\n",
    "                convs.append(nn.Conv2d(last_channel, out_channel, 1))\n",
    "                bns.append(nn.BatchNorm2d(out_channel))\n",
    "                last_channel = out_channel\n",
    "            self.conv_blocks.append(convs)\n",
    "            self.bn_blocks.append(bns)\n",
    "\n",
    "    def forward(self, xyz, points):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            xyz: input points position data, [B, C, N]\n",
    "            points: input points data, [B, D, N]\n",
    "        Return:\n",
    "            new_xyz: sampled points position data, [B, C, S]\n",
    "            new_points_concat: sample points feature data, [B, D', S]\n",
    "        \"\"\"\n",
    "        xyz = xyz.permute(0, 2, 1)\n",
    "        if points is not None:\n",
    "            points = points.permute(0, 2, 1)\n",
    "\n",
    "        B, N, C = xyz.shape\n",
    "        S = self.npoint\n",
    "        new_xyz = index_points(xyz, farthest_point_sample(xyz, S))\n",
    "        new_points_list = []\n",
    "        for i, radius in enumerate(self.radius_list):\n",
    "            K = self.nsample_list[i]\n",
    "            group_idx = query_ball_point(radius, K, xyz, new_xyz)\n",
    "            grouped_xyz = index_points(xyz, group_idx)\n",
    "            grouped_xyz -= new_xyz.view(B, S, 1, C)\n",
    "            if points is not None:\n",
    "                grouped_points = index_points(points, group_idx)\n",
    "                grouped_points = torch.cat([grouped_points, grouped_xyz], dim=-1)\n",
    "            else:\n",
    "                grouped_points = grouped_xyz\n",
    "\n",
    "            grouped_points = grouped_points.permute(0, 3, 2, 1)  # [B, D, K, S]\n",
    "            for j in range(len(self.conv_blocks[i])):\n",
    "                conv = self.conv_blocks[i][j]\n",
    "                bn = self.bn_blocks[i][j]\n",
    "                grouped_points =  F.relu(bn(conv(grouped_points)))\n",
    "            new_points = torch.max(grouped_points, 2)[0]  # [B, D', S]\n",
    "            new_points_list.append(new_points)\n",
    "\n",
    "        new_xyz = new_xyz.permute(0, 2, 1)\n",
    "        new_points_concat = torch.cat(new_points_list, dim=1)\n",
    "        return new_xyz, new_points_concat\n",
    "\n",
    "\n",
    "class PointNetFeaturePropagation(nn.Module):\n",
    "    def __init__(self, in_channel, mlp):\n",
    "        super(PointNetFeaturePropagation, self).__init__()\n",
    "        self.mlp_convs = nn.ModuleList()\n",
    "        self.mlp_bns = nn.ModuleList()\n",
    "        last_channel = in_channel\n",
    "        for out_channel in mlp:\n",
    "            self.mlp_convs.append(nn.Conv1d(last_channel, out_channel, 1))\n",
    "            self.mlp_bns.append(nn.BatchNorm1d(out_channel))\n",
    "            last_channel = out_channel\n",
    "\n",
    "    def forward(self, xyz1, xyz2, points1, points2):\n",
    "        \"\"\"\n",
    "        Input:\n",
    "            xyz1: input points position data, [B, C, N]\n",
    "            xyz2: sampled input points position data, [B, C, S]\n",
    "            points1: input points data, [B, D, N]\n",
    "            points2: input points data, [B, D, S]\n",
    "        Return:\n",
    "            new_points: upsampled points data, [B, D', N]\n",
    "        \"\"\"\n",
    "        xyz1 = xyz1.permute(0, 2, 1)\n",
    "        xyz2 = xyz2.permute(0, 2, 1)\n",
    "\n",
    "        points2 = points2.permute(0, 2, 1)\n",
    "        B, N, C = xyz1.shape\n",
    "        _, S, _ = xyz2.shape\n",
    "\n",
    "        if S == 1:\n",
    "            interpolated_points = points2.repeat(1, N, 1)\n",
    "        else:\n",
    "            dists = square_distance(xyz1, xyz2)\n",
    "            dists, idx = dists.sort(dim=-1)\n",
    "            dists, idx = dists[:, :, :3], idx[:, :, :3]  # [B, N, 3]\n",
    "\n",
    "            dist_recip = 1.0 / (dists + 1e-8)\n",
    "            norm = torch.sum(dist_recip, dim=2, keepdim=True)\n",
    "            weight = dist_recip / norm\n",
    "            interpolated_points = torch.sum(index_points(points2, idx) * weight.view(B, N, 3, 1), dim=2)\n",
    "\n",
    "        if points1 is not None:\n",
    "            points1 = points1.permute(0, 2, 1)\n",
    "            new_points = torch.cat([points1, interpolated_points], dim=-1)\n",
    "        else:\n",
    "            new_points = interpolated_points\n",
    "\n",
    "        new_points = new_points.permute(0, 2, 1)\n",
    "        for i, conv in enumerate(self.mlp_convs):\n",
    "            bn = self.mlp_bns[i]\n",
    "            new_points = F.relu(bn(conv(new_points)))\n",
    "        return new_points\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "executionInfo": {
     "elapsed": 1042,
     "status": "ok",
     "timestamp": 1674079810838,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "6A7mwh0Y2MxG",
    "tags": []
   },
   "outputs": [],
   "source": [
    "class get_model(nn.Module):\n",
    "    def __init__(self,num_class,normal_channel=True):\n",
    "        super(get_model, self).__init__()\n",
    "        in_channel = 3 if normal_channel else 0\n",
    "        self.normal_channel = normal_channel\n",
    "        self.sa1 = PointNetSetAbstractionMsg(512, [0.1, 0.2, 0.4], [16, 32, 128], in_channel,[[32, 32, 64], [64, 64, 128], [64, 96, 128]])\n",
    "        self.sa2 = PointNetSetAbstractionMsg(128, [0.2, 0.4, 0.8], [32, 64, 128], 320,[[64, 64, 128], [128, 128, 256], [128, 128, 256]])\n",
    "        self.sa3 = PointNetSetAbstraction(None, None, None, 640 + 3, [256, 512, 1024], True)\n",
    "        self.fc1 = nn.Linear(1024, 512)\n",
    "        self.bn1 = nn.BatchNorm1d(512)\n",
    "        self.drop1 = nn.Dropout(0.4)\n",
    "        self.fc2 = nn.Linear(512, 256)\n",
    "        self.bn2 = nn.BatchNorm1d(256)\n",
    "        self.drop2 = nn.Dropout(0.5)\n",
    "        self.fc3 = nn.Linear(256, num_class)\n",
    "\n",
    "    def forward(self, xyz):\n",
    "        B, _, _ = xyz.shape\n",
    "        if self.normal_channel:\n",
    "            norm = xyz[:, 3:, :]\n",
    "            xyz = xyz[:, :3, :]\n",
    "        else:\n",
    "            norm = None\n",
    "        l1_xyz, l1_points = self.sa1(xyz, norm)\n",
    "        l2_xyz, l2_points = self.sa2(l1_xyz, l1_points)\n",
    "        l3_xyz, l3_points = self.sa3(l2_xyz, l2_points)\n",
    "        x = l3_points.view(B, 1024)\n",
    "        x = self.drop1(F.relu(self.bn1(self.fc1(x))))\n",
    "        x = self.drop2(F.relu(self.bn2(self.fc2(x))))\n",
    "        x = self.fc3(x)\n",
    "        x = F.log_softmax(x, -1)\n",
    "\n",
    "\n",
    "        return x,l3_points\n",
    "\n",
    "\n",
    "class get_loss(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(get_loss, self).__init__()\n",
    "\n",
    "    def forward(self, pred, target, trans_feat):\n",
    "        total_loss = F.nll_loss(pred, target)\n",
    "\n",
    "        return total_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [],
   "source": [
    "CUDA_LAUNCH_BLOCKING=\"1\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GnM8yu2G2XVF"
   },
   "source": [
    "# Training\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.cuda.empty_cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 290,
     "status": "ok",
     "timestamp": 1674079813770,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "v2_wk24U2cc-",
    "outputId": "cc19c979-2e04-4161-cafd-b53c425b0a71"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda:0\n"
     ]
    }
   ],
   "source": [
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cpu\n"
     ]
    }
   ],
   "source": [
    "device = torch.device('cpu')\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [],
   "source": [
    "def inplace_relu(m):\n",
    "    classname = m.__class__.__name__\n",
    "    if classname.find('ReLU') != -1:\n",
    "        m.inplace=True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [],
   "source": [
    "classifier = get_model(6, normal_channel=False)\n",
    "criterion = get_loss()\n",
    "classifier.apply(inplace_relu)\n",
    "\n",
    "classifier = classifier.to(device)\n",
    "criterion = criterion.to(device)\n",
    "\n",
    "optimizer = torch.optim.Adam(classifier.parameters(), lr=0.001)\n",
    "\n",
    "scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=20, gamma=0.7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test(model, loader, num_class=6):\n",
    "    mean_correct = []\n",
    "    class_acc = np.zeros((num_class, 3))\n",
    "    classifier = model.eval()\n",
    "\n",
    "    for j, data in tqdm(enumerate(loader), total=len(loader)):\n",
    "\n",
    "        target = data[:,1,3]\n",
    "        target = target.type(torch.LongTensor) \n",
    "        points, target = data[:,:,0:3].to(device).float(), target.to(device)\n",
    "\n",
    "        points = points.transpose(2, 1)\n",
    "        pred, _ = classifier(points)\n",
    "        pred_choice = pred.data.max(1)[1]\n",
    "\n",
    "        for cat in np.unique(target.cpu()):\n",
    "            classacc = pred_choice[target == cat].eq(target[target == cat].long().data).cpu().sum()\n",
    "            class_acc[cat, 0] += classacc.item() / float(points[target == cat].size()[0])\n",
    "            class_acc[cat, 1] += 1\n",
    "\n",
    "        correct = pred_choice.eq(target.long().data).cpu().sum()\n",
    "        mean_correct.append(correct.item() / float(points.size()[0]))\n",
    "\n",
    "    class_acc[:, 2] = class_acc[:, 0] / class_acc[:, 1]\n",
    "    class_acc = np.mean(class_acc[:, 2])\n",
    "    instance_acc = np.mean(mean_correct)\n",
    "\n",
    "    return instance_acc, class_acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "executionInfo": {
     "elapsed": 288,
     "status": "ok",
     "timestamp": 1674079819088,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "XLE6uDz2VfGd"
   },
   "outputs": [],
   "source": [
    "def train(classifier, criterion, train_loader, optimizer, scheduler, val_loader=None, start_epoch = 0, epochs=20, save=True):\n",
    "    \n",
    "    global_epoch = 0\n",
    "    global_step = 0\n",
    "    best_instance_acc = 0.0\n",
    "    best_class_acc = 0.0\n",
    "    \n",
    "    for epoch in range(start_epoch, epochs):\n",
    "        print('Epoch %d (%d/%s):' % (global_epoch + 1, epoch + 1, epochs))\n",
    "        mean_correct = []\n",
    "        classifier = classifier.train()\n",
    "\n",
    "        scheduler.step()\n",
    "        for batch_id, data in tqdm(enumerate(train_loader, 0), total=len(train_loader), smoothing=0.9):\n",
    "            optimizer.zero_grad()\n",
    "            \n",
    "            target = data[:,1,3]\n",
    "            target = target.type(torch.LongTensor) \n",
    "            points, target = data[:,:,0:3].to(device).float(), target.to(device)\n",
    "            points = points.transpose(2, 1)\n",
    "\n",
    "            pred, trans_feat = classifier(points)\n",
    "            loss = criterion(pred, target.long(), trans_feat)\n",
    "            pred_choice = pred.data.max(1)[1]\n",
    "\n",
    "            correct = pred_choice.eq(target.long().data).cpu().sum()\n",
    "            mean_correct.append(correct.item() / float(points.size()[0]))\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            global_step += 1\n",
    "\n",
    "        train_instance_acc = np.mean(mean_correct)\n",
    "        print('Train Instance Accuracy: %f' % train_instance_acc)\n",
    "\n",
    "        with torch.no_grad():\n",
    "            instance_acc, class_acc = test(classifier.eval(), val_loader, num_class=6)\n",
    "\n",
    "            if (instance_acc >= best_instance_acc):\n",
    "                best_instance_acc = instance_acc\n",
    "                best_epoch = epoch + 1\n",
    "\n",
    "            if (class_acc >= best_class_acc):\n",
    "                best_class_acc = class_acc\n",
    "            print('Test Instance Accuracy: %f, Class Accuracy: %f' % (instance_acc, class_acc))\n",
    "            print('Best Instance Accuracy: %f, Class Accuracy: %f' % (best_instance_acc, best_class_acc))\n",
    "\n",
    "            if (instance_acc >= best_instance_acc):\n",
    "                print('Save model...')\n",
    "                savepath = 'best_model_p2_xyz_sd.pth'\n",
    "                print('Saving at %s' % savepath)\n",
    "                state = {\n",
    "                    'epoch': best_epoch,\n",
    "                    'instance_acc': instance_acc,\n",
    "                    'class_acc': class_acc,\n",
    "                    'model_state_dict': classifier.state_dict(),\n",
    "                    'optimizer_state_dict': optimizer.state_dict(),\n",
    "                }\n",
    "                torch.save(state, savepath)\n",
    "            global_epoch += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.environ[\"PYTORCH_CUDA_ALLOC_CONF\"] = \"max_split_size_mb:512\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5476095,
     "status": "ok",
     "timestamp": 1674085297893,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "EOYHr6IW2Qsh",
    "outputId": "703e8b98-e753-43bc-c3b8-ea719bf8ffff"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda\\Anaconda_py3.9\\lib\\site-packages\\torch\\optim\\lr_scheduler.py:138: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule. See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
      "  warnings.warn(\"Detected call of `lr_scheduler.step()` before `optimizer.step()`. \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 (1/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:39<00:00,  2.27it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.618943\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:30<00:00,  2.47it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.732456, Class Accuracy: 0.587475\n",
      "Best Instance Accuracy: 0.732456, Class Accuracy: 0.587475\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 2 (2/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:41<00:00,  2.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.726872\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:29<00:00,  2.57it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.713816, Class Accuracy: 0.631939\n",
      "Best Instance Accuracy: 0.732456, Class Accuracy: 0.631939\n",
      "Epoch 3 (3/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:41<00:00,  2.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.755507\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:32<00:00,  2.37it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.907895, Class Accuracy: 0.715398\n",
      "Best Instance Accuracy: 0.907895, Class Accuracy: 0.715398\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 4 (4/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:42<00:00,  2.22it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.780837\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:30<00:00,  2.50it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.881579, Class Accuracy: 0.672607\n",
      "Best Instance Accuracy: 0.907895, Class Accuracy: 0.715398\n",
      "Epoch 5 (5/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:41<00:00,  2.23it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.813877\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:29<00:00,  2.54it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.894737, Class Accuracy: 0.684312\n",
      "Best Instance Accuracy: 0.907895, Class Accuracy: 0.715398\n",
      "Epoch 6 (6/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:37<00:00,  2.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.831498\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:28<00:00,  2.71it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.914474, Class Accuracy: 0.734598\n",
      "Best Instance Accuracy: 0.914474, Class Accuracy: 0.734598\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 7 (7/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:33<00:00,  2.43it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.825991\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:30<00:00,  2.52it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.917763, Class Accuracy: 0.744795\n",
      "Best Instance Accuracy: 0.917763, Class Accuracy: 0.744795\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 8 (8/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:41<00:00,  2.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.851322\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:30<00:00,  2.49it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.940789, Class Accuracy: 0.774248\n",
      "Best Instance Accuracy: 0.940789, Class Accuracy: 0.774248\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 9 (9/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:36<00:00,  2.35it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.850220\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:29<00:00,  2.59it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.867325, Class Accuracy: 0.793876\n",
      "Best Instance Accuracy: 0.940789, Class Accuracy: 0.793876\n",
      "Epoch 10 (10/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:41<00:00,  2.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.852423\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:28<00:00,  2.64it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.944079, Class Accuracy: 0.767415\n",
      "Best Instance Accuracy: 0.944079, Class Accuracy: 0.793876\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 11 (11/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:35<00:00,  2.38it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.878855\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:31<00:00,  2.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.894737, Class Accuracy: 0.752640\n",
      "Best Instance Accuracy: 0.944079, Class Accuracy: 0.793876\n",
      "Epoch 12 (12/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:36<00:00,  2.36it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.809471\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:29<00:00,  2.61it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.850877, Class Accuracy: 0.714909\n",
      "Best Instance Accuracy: 0.944079, Class Accuracy: 0.793876\n",
      "Epoch 13 (13/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:34<00:00,  2.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.863436\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:28<00:00,  2.65it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.963816, Class Accuracy: 0.801132\n",
      "Best Instance Accuracy: 0.963816, Class Accuracy: 0.801132\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 14 (14/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 227/227 [01:33<00:00,  2.43it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Instance Accuracy: 0.841410\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████████████████████████████████████████████████| 76/76 [00:28<00:00,  2.66it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Instance Accuracy: 0.973684, Class Accuracy: 0.810094\n",
      "Best Instance Accuracy: 0.973684, Class Accuracy: 0.810094\n",
      "Save model...\n",
      "Saving at best_model_p2_xyz_sd.pth\n",
      "Epoch 15 (15/20):\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|███▌                                                                             | 10/227 [00:04<01:30,  2.39it/s]\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_18340\\4000209220.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mtrain\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mclassifier\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrain_loader\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mscheduler\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvalid_loader\u001b[0m\u001b[1;33m,\u001b[0m  \u001b[0msave\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_18340\\181406597.py\u001b[0m in \u001b[0;36mtrain\u001b[1;34m(classifier, criterion, train_loader, optimizer, scheduler, val_loader, start_epoch, epochs, save)\u001b[0m\n\u001b[0;32m     20\u001b[0m             \u001b[0mpoints\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpoints\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtranspose\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     21\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 22\u001b[1;33m             \u001b[0mpred\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrans_feat\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mclassifier\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpoints\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     23\u001b[0m             \u001b[0mloss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpred\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtarget\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlong\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrans_feat\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     24\u001b[0m             \u001b[0mpred_choice\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mpred\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmax\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mD:\\Anaconda\\Anaconda_py3.9\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m   1192\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[0;32m   1193\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[1;32m-> 1194\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1195\u001b[0m         \u001b[1;31m# Do not call functions when jit is used\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1196\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_18340\\67312410.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, xyz)\u001b[0m\n\u001b[0;32m     23\u001b[0m             \u001b[0mnorm\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     24\u001b[0m         \u001b[0ml1_xyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ml1_points\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msa1\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnorm\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 25\u001b[1;33m         \u001b[0ml2_xyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ml2_points\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msa2\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ml1_xyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ml1_points\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     26\u001b[0m         \u001b[0ml3_xyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ml3_points\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msa3\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0ml2_xyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ml2_points\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     27\u001b[0m         \u001b[0mx\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0ml3_points\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mview\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1024\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mD:\\Anaconda\\Anaconda_py3.9\\lib\\site-packages\\torch\\nn\\modules\\module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[1;34m(self, *input, **kwargs)\u001b[0m\n\u001b[0;32m   1192\u001b[0m         if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks or _global_backward_hooks\n\u001b[0;32m   1193\u001b[0m                 or _global_forward_hooks or _global_forward_pre_hooks):\n\u001b[1;32m-> 1194\u001b[1;33m             \u001b[1;32mreturn\u001b[0m \u001b[0mforward_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0minput\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1195\u001b[0m         \u001b[1;31m# Do not call functions when jit is used\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1196\u001b[0m         \u001b[0mfull_backward_hooks\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnon_full_backward_hooks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_18340\\4109883452.py\u001b[0m in \u001b[0;36mforward\u001b[1;34m(self, xyz, points)\u001b[0m\n\u001b[0;32m    231\u001b[0m         \u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mN\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mC\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mxyz\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    232\u001b[0m         \u001b[0mS\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnpoint\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 233\u001b[1;33m         \u001b[0mnew_xyz\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mindex_points\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfarthest_point_sample\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxyz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mS\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    234\u001b[0m         \u001b[0mnew_points_list\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    235\u001b[0m         \u001b[1;32mfor\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mradius\u001b[0m \u001b[1;32min\u001b[0m \u001b[0menumerate\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mradius_list\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp\\ipykernel_18340\\4109883452.py\u001b[0m in \u001b[0;36mfarthest_point_sample\u001b[1;34m(xyz, npoint)\u001b[0m\n\u001b[0;32m     70\u001b[0m     \u001b[0mbatch_indices\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0marange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlong\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mto\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdevice\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     71\u001b[0m     \u001b[1;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnpoint\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 72\u001b[1;33m         \u001b[0mcentroids\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mi\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfarthest\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     73\u001b[0m         \u001b[0mcentroid\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mxyz\u001b[0m\u001b[1;33m[\u001b[0m\u001b[0mbatch_indices\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfarthest\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m:\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mview\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mB\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m3\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     74\u001b[0m         \u001b[0mdist\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msum\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mxyz\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mcentroid\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m**\u001b[0m \u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "train(classifier, criterion, train_loader, optimizer, scheduler, valid_loader,  save=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rdIlrnXh2UX9"
   },
   "source": [
    "# Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "executionInfo": {
     "elapsed": 269,
     "status": "ok",
     "timestamp": 1674085428035,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "mNefMkzH2Wvw"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "id": "h2u-0bwC2X_Y"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 122,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "checkpoint = torch.load('best_model_p2_xyz_sd.pth')\n",
    "classifier.load_state_dict(checkpoint['model_state_dict'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "def test(model, test_loader, num_class=6, vote_num=1):\n",
    "    mean_correct = []\n",
    "    classifier = model.eval()\n",
    "    class_acc = np.zeros((num_class, 3))\n",
    "    all_preds = []\n",
    "    all_labels = []\n",
    "\n",
    "    for j, data in tqdm(enumerate(test_loader), total=len(test_loader)):\n",
    "        \n",
    "        target = data[:,1,3]\n",
    "        target = target.type(torch.LongTensor) \n",
    "        points, target = data[:,:,0:3].to(device).float(), target.to(device)\n",
    "        points = points.transpose(2, 1)\n",
    "        vote_pool = torch.zeros(target.size()[0], num_class).cuda()\n",
    "\n",
    "        for _ in range(vote_num):\n",
    "            pred, _ = classifier(points)\n",
    "            vote_pool += pred\n",
    "        pred = vote_pool / vote_num\n",
    "        pred_choice = pred.data.max(1)[1]\n",
    "        all_preds += list(pred_choice.cpu().numpy())\n",
    "        all_labels += list(target.cpu().numpy())\n",
    "\n",
    "        for cat in np.unique(target.cpu()):\n",
    "            classacc = pred_choice[target == cat].eq(target[target == cat].long().data).cpu().sum()\n",
    "            class_acc[cat, 0] += classacc.item() / float(points[target == cat].size()[0])\n",
    "            class_acc[cat, 1] += 1\n",
    "        correct = pred_choice.eq(target.long().data).cpu().sum()\n",
    "        mean_correct.append(correct.item() / float(points.size()[0]))\n",
    "\n",
    "    \n",
    "    for i in range(num_class):\n",
    "        if class_acc[i,1]>0:\n",
    "            class_acc[i, 2] = class_acc[i, 0] / class_acc[i, 1]\n",
    "    class_acc = np.mean(class_acc[:, 2])\n",
    "    instance_acc = np.mean(mean_correct)\n",
    "    return instance_acc, class_acc, all_preds, all_labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████| 153/153 [01:00<00:00,  2.55it/s]\n"
     ]
    }
   ],
   "source": [
    "ins_acc, cla_acc, all_preds, all_labels = test(classifier.eval(), test_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.4261281988828099\n",
      "0.5049019607843137\n"
     ]
    }
   ],
   "source": [
    "print(cla_acc)\n",
    "print(ins_acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 376,
     "status": "ok",
     "timestamp": 1674088910585,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "LvqVmt0i2ds3",
    "outputId": "2f890904-33c1-49e0-c5a9-7ba544a7cef6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[226 146   2  55 169  12]\n",
      "[100 138   0   4 310  58]\n",
      "[[ 94  87   0   0  44   1]\n",
      " [  5  42   0   0  99   0]\n",
      " [  0   0   0   0   1   1]\n",
      " [  1   8   0   3   8  35]\n",
      " [  0   1   0   1 157  10]\n",
      " [  0   0   0   0   1  11]]\n",
      "0.5032786885245901\n"
     ]
    }
   ],
   "source": [
    "cm = confusion_matrix(all_labels, all_preds);\n",
    "cm\n",
    "print(np.bincount(all_labels))\n",
    "print(np.bincount(all_preds))\n",
    "print(cm)\n",
    "\n",
    "correct_rate = np.diagonal(cm).sum()/cm.sum()\n",
    "print(correct_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 245,
     "status": "ok",
     "timestamp": 1674086880589,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "jQryCKH4R1FW",
    "outputId": "528acf69-4698-4cea-be2a-ded3b34f69d9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'wall & balcony & ..': 0,\n",
       " 'window & blinds': 1,\n",
       " 'door': 2,\n",
       " 'molding & decoration': 3,\n",
       " 'ground surface & terrian': 4,\n",
       " 'roof & others': 5}"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "classes = {'wall & balcony & ..': 0,\n",
    "           'window & blinds': 1,\n",
    "           'door': 2,\n",
    "           'molding & decoration': 3,\n",
    "           'ground surface & terrian': 4,\n",
    "           'roof & others': 5\n",
    "           }\n",
    "classes\n",
    "cla_test = {'wall & balcony & ..': 0,\n",
    "           'window & blinds': 1,\n",
    "           'door': 2,\n",
    "           'molding & decoration': 3,\n",
    "           'ground surface & terrian': 4,\n",
    "           'roof & others': 5\n",
    "           }\n",
    "cla_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "metadata": {
    "executionInfo": {
     "elapsed": 235,
     "status": "ok",
     "timestamp": 1674086267324,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "MvIkJrRI2g8L"
   },
   "outputs": [],
   "source": [
    "import itertools\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# function from https://deeplizard.com/learn/video/0LhiS6yu2qQ\n",
    "def plot_confusion_matrix(cm, classes, normalize=False, title='Confusion matrix', cmap=plt.cm.Blues):\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "        print(\"Normalized confusion matrix\")\n",
    "    else:\n",
    "        print('Confusion matrix, without normalization')\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=45)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt), horizontalalignment=\"center\", color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 652
    },
    "executionInfo": {
     "elapsed": 830,
     "status": "ok",
     "timestamp": 1674086917326,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "YJeGeH212i6h",
    "outputId": "72dd3fa4-5411-44bb-c876-8f1e7cb6d839"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Normalized confusion matrix\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,8))\n",
    "plot_confusion_matrix(cm, list(cla_test.keys()), normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 580
    },
    "executionInfo": {
     "elapsed": 1205,
     "status": "ok",
     "timestamp": 1674086946985,
     "user": {
      "displayName": "SUKIDA Tan",
      "userId": "04234593673225449814"
     },
     "user_tz": -60
    },
    "id": "9FeHmyzh2mZq",
    "outputId": "8cb5c54e-31e6-448c-c9c5-5b8b2c48c0bf"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Confusion matrix, without normalization\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 800x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,8))\n",
    "plot_confusion_matrix(cm, list(cla_test.keys()), normalize=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "authorship_tag": "ABX9TyOps7S6O6rs3xgWiavpNy6T",
   "collapsed_sections": [
    "be8RFvYTsfYE",
    "sSRH6VAGf7Xw",
    "N0FRlJOh2EuO"
   ],
   "provenance": [],
   "toc_visible": true
  },
  "gpuClass": "standard",
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}