nfvae.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from me_utils import *\n",
    "from scipy.stats import wasserstein_distance\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "# import keras\n",
    "# from tensorflow import keras\n",
    "import pandas as pd\n",
    "from sklearn.neighbors import KernelDensity\n",
    "\n",
    "# from tensorflow.keras.layers import Lambda, Input, Dense, Add, Activation, Concatenate\n",
    "# from tensorflow.keras.models import Model\n",
    "# from tensorflow.keras import initializers\n",
    "# from tensorflow.keras import backend as K\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import glob\n",
    "from matplotlib.cm import get_cmap\n",
    "from cycler import cycler\n",
    "import sklearn\n",
    "from sklearn.utils import shuffle\n",
    "\n",
    "import torch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "nFeat = 4\n",
    "trainsize = 499889\n",
    "intermediate_dim = 128\n",
    "encoder_dim = 128\n",
    "batch_size = 1024\n",
    "latent_dim = 20\n",
    "epochs = 240\n",
    "eluvar = np.sqrt(1.55/intermediate_dim)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def sampling(args):\n",
    "#     \"\"\"Reparameterization trick by sampling fr an isotropic unit Gaussian.\n",
    "#     # Arguments:\n",
    "#         args (tensor): mean and log of variance of Q(z|X)\n",
    "#     # Returns:\n",
    "#         z (tensor): sampled latent vector\n",
    "#     \"\"\"\n",
    "\n",
    "#     z_mean, z_log_var = args\n",
    "#     batch = K.shape(z_mean)[0]\n",
    "#     dim = K.int_shape(z_mean)[1]\n",
    "#     # by default, random_normal has mean=0 and std=1.0\n",
    "#     epsilon = K.random_normal(shape=(batch, dim))\n",
    "#     return z_mean + K.exp(0.5 * z_log_var) * epsilon"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# load model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def create_encoder(X,y):\n",
    "# \tinputs = Concatenate()([X, y])\n",
    "# \tx1 = Dense(encoder_dim, activation='elu', kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(inputs)\n",
    "# \tx2 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x1)\n",
    "# \tx2 = Activation('elu')(x2)\n",
    "# \tx3 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x2)\n",
    "# \tsc1 = Add()([x1,x3])\n",
    "# \tx3 = Activation('elu')(sc1)\n",
    "# \tx4 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x3)\n",
    "# \tsc2 = Add()([x2,x4])\n",
    "# \tx4 = Activation('elu')(sc2)\n",
    "# \tx5 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x4)\n",
    "# \tsc3 = Add()([x3,x5])\n",
    "# \tx5 = Activation('elu')(sc3)\n",
    "# \tx6 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x5)\n",
    "# \tsc4 = Add()([x4,x6])\n",
    "# \tx6 = Activation('elu')(sc4)\n",
    "# \tz_mean = Dense(latent_dim, name='z_mean')(x6)\n",
    "# \tz_log_var = Dense(latent_dim, name='z_log_var')(x6)\n",
    "\n",
    "# \t# use reparameterization trick to push the sampling out as input\n",
    "# \t# note that \"output_shape\" isn't necessary with the TensorFlow backend\n",
    "# \tz = Lambda(sampling, output_shape=(latent_dim,), name='z')([z_mean, z_log_var])\n",
    "\n",
    "# \treturn Model([X, y], [z_mean, z_log_var, z], name='encoder')\n",
    "\n",
    "# encoder = create_encoder(Input(shape=(nFeat,)), Input(shape=(1,)))\n",
    "# encoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def create_decoder(z,y):\n",
    "# \tlatent_inputs = Concatenate()([z, y])\n",
    "# \tx1 = Dense(intermediate_dim, activation='elu', kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(latent_inputs)\n",
    "# \tx2 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x1)\n",
    "# \tx2 = Activation('elu')(x2)\n",
    "# \tx3 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x2)\n",
    "# \tsc1 = Add()([x1,x3])\n",
    "# \tx3 = Activation('elu')(sc1)\n",
    "# \tx4 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x3)\n",
    "# \tsc2 = Add()([x2,x4])\n",
    "# \tx4 = Activation('elu')(sc2)\n",
    "# \tx5 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x4)\n",
    "# \tsc3 = Add()([x3,x5])\n",
    "# \tx5 = Activation('elu')(sc3)\n",
    "# \tx6 = Dense(encoder_dim, kernel_initializer=initializers.random_normal(mean=0.0, stddev=eluvar))(x5)\n",
    "# \tsc4 = Add()([x4,x6])\n",
    "# \tx6 = Activation('elu')(sc4)\n",
    "# \toutputs = Dense(nFeat+1, activation='tanh')(x6)\n",
    "\n",
    "# \treturn Model([z, y], outputs, name='decoder')\n",
    "\n",
    "# # instantiate decoder model\n",
    "# decoder = create_decoder(Input(shape=(latent_dim,)), Input(shape=(1,)))\n",
    "# decoder.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# X_input = Input(shape=(nFeat,))\n",
    "# y_input = Input(shape=(1,))\n",
    "# z_mean, z_log_var, z = encoder([X_input,y_input])\n",
    "# outputs = decoder([z,y_input])\n",
    "# cvae = Model([X_input, y_input], outputs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model1 = keras.models.load_model('/global/u2/a/agarabag/new_models/cbvae_LHCO2020_epoch240.h5')\n",
    "import torch\n",
    "model = torch.load(\"/global/u2/a/agarabag/DarkFlow/model_save/BEST_test_ConvF.pt\", map_location=torch.device('cpu'))\n",
    "# model.eval()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_of_cgen_models= [model]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load real data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "innerdata_train = np.load(\"/global/homes/a/agarabag/anomoly_studies/test_cathode/CATHODE/separated_data/innerdata_train.npy\")\n",
    "innerdata_train = innerdata_train[innerdata_train[:,nFeat+1]==0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of events in innerdata_train: 59605\n",
      "number of events in innerdata with label=0 (bkg): 59605\n"
     ]
    }
   ],
   "source": [
    "# check number of total events & bkg in innerdata_train\n",
    "\n",
    "pd_innerdata_train = pd.DataFrame(innerdata_train, columns=[\"mjj\", \"mjmin\", \"mjmax-mjmin\", \"tau21j1\", \"tau21j2\", \"label\"])\n",
    "print(\"number of events in innerdata_train:\", len(pd_innerdata_train))\n",
    "pd_innerdata_train_bkg = pd_innerdata_train[pd_innerdata_train[\"label\"]==0]\n",
    "print(\"number of events in innerdata with label=0 (bkg):\", len(pd_innerdata_train_bkg))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape of X_innerdata_train: (59605, 4)\n",
      "shape of y_innerdata_train: (59605,)\n"
     ]
    }
   ],
   "source": [
    "X_innerdata_train = innerdata_train[:,1:nFeat+1]\n",
    "y_innerdata_train = innerdata_train[:,0]\n",
    "\n",
    "print(\"shape of X_innerdata_train:\", X_innerdata_train.shape)\n",
    "print(\"shape of y_innerdata_train:\", y_innerdata_train.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_features = [\"mjmin\", \"mjmax_minus_mjmin\", \"tau21j1\", \"tau21j2\"]\n",
    "condition_features = [\"mjj\"]\n",
    "\n",
    "nFeat = len(train_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create scaler\n",
    "\n",
    "outerdata_train = np.load(\"/global/homes/a/agarabag/anomoly_studies/test_cathode/CATHODE/separated_data/outerdata_train.npy\")\n",
    "outerdata_test = np.load(\"/global/homes/a/agarabag/anomoly_studies/test_cathode/CATHODE/separated_data/outerdata_test.npy\")\n",
    "outerdata_train = outerdata_train[outerdata_train[:,nFeat+1]==0]\n",
    "outerdata_test = outerdata_test[outerdata_test[:,nFeat+1]==0]\n",
    "\n",
    "X_data_train_unscaled = outerdata_train[:,1:nFeat+1]\n",
    "y_data_train_unscaled = outerdata_train[:,0]\n",
    "\n",
    "X_data_test_unscaled = outerdata_test[:,1:nFeat+1]\n",
    "y_data_test_unscaled = outerdata_test[:,0]\n",
    "\n",
    "X_data_unscaled = np.concatenate((X_data_train_unscaled, X_data_test_unscaled), axis=0)\n",
    "X_data = np.copy(X_data_unscaled)\n",
    "y_data = np.concatenate((y_data_train_unscaled, y_data_test_unscaled), axis=0)\n",
    "\n",
    "# scaler = MinMaxScaler((0,1)).fit(X_data)\n",
    "# scaler_mjj = MinMaxScaler((0,1)).fit(y_data.reshape(-1,1))\n",
    "x_max = np.empty(nFeat)\n",
    "for i in range(0,X_data.shape[1]):\n",
    "\tx_max[i] = np.max(np.abs(X_data[:,i]))\n",
    "\tif np.abs(x_max[i]) > 0: \n",
    "\t\tX_data[:,i] = X_data[:,i]/x_max[i]\n",
    "\telse:\n",
    "\t\tpass\n",
    "\n",
    "cond_max = np.max(np.abs(y_data))\n",
    "if np.abs(cond_max) > 0:\n",
    "    y_data = y_data/cond_max\n",
    "else:\n",
    "    pass"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# generate mjj distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def quick_logit(x):\n",
    "    x_norm = (x-min(x))/(max(x)-min(x))\n",
    "    x_norm = x_norm[(x_norm != 0) & (x_norm != 1)]\n",
    "    logit = np.log(x_norm/(1-x_norm))\n",
    "    logit = logit[~np.isnan(logit)]\n",
    "    return logit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def logit_transform_inverse(data, datamax, datamin):\n",
    "    \n",
    "    dataout = (datamin + datamax*np.exp(data))/(1 + np.exp(data))\n",
    "\n",
    "    return dataout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(59605,)\n",
      "shape of input mjj values (400000, 1)\n"
     ]
    }
   ],
   "source": [
    "# fitting and sampling KDE\n",
    "\n",
    "n_samples = 400000\n",
    "#n_samples = 1200000\n",
    "KDE_bandwidth = 0.01\n",
    "\n",
    "print(y_innerdata_train.shape)\n",
    "mjj_logit = quick_logit(y_innerdata_train)\n",
    "train_mjj_vals = logit_transform_inverse(KernelDensity(\n",
    "    bandwidth=KDE_bandwidth, kernel='gaussian').fit(\n",
    "        mjj_logit.reshape(-1, 1)).sample(n_samples),\n",
    "                                            max(y_innerdata_train).item(),\n",
    "                                            min(y_innerdata_train).item())\n",
    "\n",
    "print(\"shape of input mjj values\", train_mjj_vals.shape) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# generate data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# means_df = np.load('/global/u2/a/agarabag/DarkFlow/model_save/latent_std.npy')\n",
    "# mean = means_df.values\n",
    "# filename = '/global/u2/a/agarabag/new_models/LHCO2020/test/cbvae_LHCO2020_latent_std_20d_e-6.csv'\n",
    "# stds_df = pd.read_csv(filename, sep=' ', header=None)\n",
    "# std = stds_df.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# means_df.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "from torch.autograd import Variable\n",
    "from torch.utils.data import DataLoader\n",
    "import torchvision.transforms as transforms\n",
    "import torchvision.datasets\n",
    "from sklearn.utils import shuffle\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "# from pickle import dump\n",
    "# import numpy as np\n",
    "# import h5py\n",
    "\n",
    "import flows as flows\n",
    "from layers import MaskedConv2d, MaskedLinear, CNN_Flow_Layer, Dilation_Block\n",
    "\n",
    "\n",
    "\n",
    "##################  DEFINE MODEL  #####################\n",
    "class ConvNet(nn.Module):\n",
    "        def __init__(self):\n",
    "            super(ConvNet, self).__init__()\n",
    "\n",
    "            self.latent_dim = 15\n",
    "            self.test_mode = 0\n",
    "\n",
    "  \n",
    "            self.batch_size = 1024\n",
    "  \n",
    "                \n",
    "            self.q_z_output_dim = 16\n",
    "            self.z_size = self.latent_dim\n",
    "    \n",
    "            # self.q_z_nn = nn.Sequential(\n",
    "            #       nn.Conv2d(1, 32, kernel_size=4, stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(32),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.Conv2d(32, 16, kernel_size=5, stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(16),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.Conv2d(16, 8, kernel_size=7, stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(8),\n",
    "            #       nn.ReLU()\n",
    "            #       ) \n",
    "            #\n",
    "            # self.mult_bn1 = nn.BatchNorm1d(161) \n",
    "            # self.dense1 = nn.Linear(161,self.q_z_output_dim) #161 - full(13+3)+met, 41-4LJ, 216 - full(13+4)+met+mult\n",
    "            # self.dnn_bn1 = nn.BatchNorm1d(self.q_z_output_dim)\n",
    "            self.q_z_mean = nn.Linear(self.q_z_output_dim, self.latent_dim)\n",
    "            self.q_z_logvar = nn.Linear(self.q_z_output_dim, self.latent_dim)\n",
    "            #\n",
    "            # self.dense3 = nn.Linear(self.latent_dim, self.q_z_output_dim)\n",
    "            # self.dnn_bn3 = nn.BatchNorm1d(self.q_z_output_dim)\n",
    "            # self.dense4 = nn.Linear(self.q_z_output_dim, 161)\n",
    "            # self.dnn_bn4 = nn.BatchNorm1d(161)\n",
    "            #\n",
    "            # self.p_x_nn = nn.Sequential(\n",
    "            #       nn.ConvTranspose2d(8, 16, kernel_size=(7,1), stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(16),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.ConvTranspose2d(16, 32, kernel_size=(5,1), stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(32),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.ConvTranspose2d(32, 1, kernel_size=(3,4), stride=(1), padding=(0))\n",
    "            #       ) \n",
    "            # self.p_x_nn = nn.Sequential(\n",
    "            #       nn.ConvTranspose2d(64, 64, kernel_size=4, stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(64),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.ConvTranspose2d(64, 32, kernel_size=4, stride=(1), padding=(0)),\n",
    "            #       nn.BatchNorm2d(32),\n",
    "            #       nn.ReLU(),\n",
    "            #       nn.ConvTranspose2d(32, 16, kernel_size=4, stride=(1), padding=(0))\n",
    "            #       ) \n",
    "            self.dnn = nn.Sequential(\n",
    "                  nn.Linear(5, 64),\n",
    "                  nn.BatchNorm1d(64),\n",
    "                  nn.ReLU(),\n",
    "                  nn.Linear(64, 32),\n",
    "                  nn.BatchNorm1d(32),\n",
    "                  nn.ReLU(),\n",
    "                  nn.Linear(32, 16),\n",
    "                  nn.BatchNorm1d(16),\n",
    "                  nn.ReLU()\n",
    "                  ) \n",
    "            self.ddnn = nn.Sequential(\n",
    "                  nn.Linear(16, 64),\n",
    "                  nn.BatchNorm1d(64),\n",
    "                  nn.ReLU(),\n",
    "                  nn.Linear(64, 16),\n",
    "                  nn.BatchNorm1d(16),\n",
    "                  nn.ReLU(),\n",
    "                  nn.Linear(16, 4),\n",
    "                  # nn.BatchNorm1d(8),\n",
    "                  # nn.ReLU()\n",
    "                  ) \n",
    "            # log-det-jacobian = 0 without flows\n",
    "            self.ldj = 0\n",
    "    \n",
    "        def encode(self, x, y):\n",
    "          \n",
    "            # print(\"TTTTTTTTT: \", x.size(), y.size())\n",
    "            out = torch.cat((x, y),axis=1)\n",
    "            # print(\"TTTTTTTTT: \", out.ndim)\n",
    "            out = self.dnn(out)\n",
    "            # print(\"TTTTTTTTT0111\")\n",
    "            # out = torch.relu(out)\n",
    "\n",
    "            mean  = self.q_z_mean(out)\n",
    "            # print(\"TTTTTTTTT1\")\n",
    "            logvar = self.q_z_logvar(out)\n",
    "            # print(\"TTTTTTTTT2\")\n",
    "            return mean, logvar\n",
    "            \n",
    "    \n",
    "        def decode(self, z, y):\n",
    "            out = torch.cat((z, y),axis=1)\n",
    "            # print(\"TTTTTTTTT: \", out.size())\n",
    "            # print(\"TTTTTTTTT: \", out.ndim)\n",
    "            out = self.ddnn(out)\n",
    "            \n",
    "            return out\n",
    "    \n",
    "        def reparameterize(self, mean, logvar):\n",
    "            z = mean + torch.randn_like(mean) * torch.exp(0.5 * logvar)\n",
    "            return z\n",
    "    \n",
    "        def forward(self, x, y):\n",
    "            mean, logvar = self.encode(x, y)\n",
    "            z = self.reparameterize(mean, logvar)\n",
    "            out = self.decode(z, y)\n",
    "            return out, mean, logvar, self.ldj, z, z\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNN_Flow(nn.Module):\n",
    "    def __init__(self, dim, cnn_layers, kernel_size, test_mode=0, use_revert=True):\n",
    "        super(CNN_Flow, self).__init__()\n",
    "\n",
    "        # prepare reversion matrix\n",
    "        self.usecuda = True\n",
    "        self.use_revert = use_revert\n",
    "        self.R = Variable(torch.from_numpy(np.flip(np.eye(dim), axis=1).copy()).float(), requires_grad=False)\n",
    "        if self.usecuda:\n",
    "            self.R = self.R.cuda()\n",
    "        \n",
    "        self.layers = nn.ModuleList()\n",
    "        for i in range(cnn_layers):\n",
    "\n",
    "            block = Dilation_Block(dim, kernel_size, test_mode)\n",
    "            self.layers.append(block)\n",
    "        \n",
    "    def forward(self, x):\n",
    "        logdetSum = 0\n",
    "        output = x\n",
    "        for i in range(len(self.layers)):\n",
    "            output, logdet = self.layers[i](output)\n",
    "            # revert the dimension of the output after each block\n",
    "            if self.use_revert:\n",
    "                z = output.mm(self.R)\n",
    "            logdetSum += logdet\n",
    "\n",
    "        return z, logdetSum"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConvFlowVAE(ConvNet):\n",
    "    \"\"\"\n",
    "    Variational auto-encoder with convolutional flows in the encoder.\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self):\n",
    "        super(ConvFlowVAE, self).__init__()\n",
    "\n",
    "        # Initialize log-det-jacobian to zero\n",
    "        self.log_det_j = 0\n",
    "        self.num_flows = 4#args.num_flows # 6 for chan1\n",
    "        self.kernel_size = 7\n",
    "\n",
    "        flow_k = CNN_Flow\n",
    "\n",
    "        # Normalizing flow layers\n",
    "        self.flow = flow_k(dim=self.latent_dim, cnn_layers=self.num_flows, kernel_size=self.kernel_size, test_mode=self.test_mode)\n",
    "\n",
    "    def forward(self, x, y):\n",
    "        self.met = y\n",
    "        self.log_det_j = 0\n",
    "\n",
    "        # mean and variance of z\n",
    "        z_mu, z_var = self.encode(x, y)\n",
    "        # sample z\n",
    "        z_0 = self.reparameterize(z_mu, z_var)\n",
    "\n",
    "        # Normalizing flows\n",
    "        z_k, logdet = self.flow(z_0)\n",
    "\n",
    "        x_decoded = self.decode(z_k, y)\n",
    "\n",
    "        return x_decoded, z_mu, z_var, self.log_det_j, z_0, z_k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# model = ConvFlowVAE()\n",
    "# model.load_state_dict(torch.load(\"/global/u2/a/agarabag/DarkFlow/model_save/BEST_test_ConvF.pt\", map_location=torch.device('cpu')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_cond_data(n_model,stats,y_samples):\n",
    "    noise = np.random.normal(0,1, [stats, 4])\n",
    "    model = ConvFlowVAE()\n",
    "    model.load_state_dict(torch.load(\"/global/u2/a/agarabag/DarkFlow/model_save/BEST_test_ConvF.pt\", map_location=torch.device('cpu')))\n",
    "    model.eval()\n",
    "    # input_test = noise.cuda().float()\n",
    "    # met_test = y_samples.cuda().float()\n",
    "    with torch.no_grad():\n",
    "        input_test = torch.tensor(noise.astype('float32'))\n",
    "        met_test = torch.tensor(y_samples.astype('float32'))\n",
    "        print(\"BBBBBBB: \", input_test.shape, met_test.shape)\n",
    "        new_events, z_mu, z_var, log_det_j, z0, zk = model(input_test, met_test)\n",
    "    \n",
    "    return new_events"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "variables_names =  train_features\n",
    "vnames = train_features\n",
    "\n",
    "plotting_ranges = [(0,0.8), (0,1), (0,1), (0,1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def fake_vs_real(gen_data,real_data,doLog,names):\n",
    "    \n",
    "    plt.figure(figsize=(20,10))\n",
    "    colors = [\"steelblue\", \"darkturquoise\"]\n",
    "    for p in range(0,nFeat):\n",
    "        R = np.linspace(plotting_ranges[p][0],plotting_ranges[p][1],51)\n",
    "        plt.subplot((nFeat+1)//2, 2, p+1)\n",
    "        for rd in real_data: b_hist_y, b_hist_x, _ = plt.hist(rd[:,p], R, color=\"black\", histtype='step', linewidth=2, label=\"Real Bkg\", density=True)        \n",
    "        for i,n in enumerate(gen_data): plt.hist(n[:,p], R, color=colors[i], histtype='stepfilled', alpha=0.5, linewidth=2, label=\"Generated Bkg %s\"%names[i], density=True)\n",
    "        # s_hist_y, s_hist_x, _ = plt.hist(sig_data[:,p], R, color=\"r\", histtype='step', linewidth=3, label=\"Signal\", density=True)\n",
    "        plt.xlabel(\"%s\"%variables_names[p])\n",
    "        if doLog == True: plt.yscale('log')\n",
    "        plt.ylabel(\"Number of Events / bin\")\n",
    "        plt.legend(fontsize=\"large\")\n",
    "        ymax_b = np.max(b_hist_y)\n",
    "        # ymax_s = np.max(s_hist_y)\n",
    "        plt.ylim(0,ymax_b*1.5)\n",
    "    plt.savefig(\"fake_gen_compare_cbvae_4var.png\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TTTTTT:  400000 (400000, 1) (400000, 1)\n",
      "BBBBBBB:  torch.Size([400000, 4]) torch.Size([400000, 1])\n"
     ]
    },
    {
     "ename": "RuntimeError",
     "evalue": "Tensor for 'out' is on CPU, Tensor for argument #1 'self' is on CPU, but expected them to be on GPU (while checking arguments for addmm)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-42-e501791dbe14>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m     \u001b[0mprint\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m\"TTTTTT: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_mjj_vals\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_mjj_vals\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_mjj_vals_scaled\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 10\u001b[0;31m     \u001b[0mfake_SR_t\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgenerate_cond_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mtrain_mjj_vals\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mtrain_mjj_vals_scaled\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     11\u001b[0m     \u001b[0;31m# fake_SR = fake_SR_t.cpu().detach().numpy()\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-39-d6bd28e5d1d3>\u001b[0m in \u001b[0;36mgenerate_cond_data\u001b[0;34m(n_model, stats, y_samples)\u001b[0m\n\u001b[1;32m     10\u001b[0m         \u001b[0mmet_test\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_samples\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mastype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'float32'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     11\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"BBBBBBB: \"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput_test\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmet_test\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 12\u001b[0;31m         \u001b[0mnew_events\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mz_mu\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mz_var\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlog_det_j\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mz0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mzk\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmet_test\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     13\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mnew_events\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/common/software/pytorch/1.8.0-gpu/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    888\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-37-e7f3fcbe38d6>\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x, y)\u001b[0m\n\u001b[1;32m     27\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     28\u001b[0m         \u001b[0;31m# Normalizing flows\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 29\u001b[0;31m         \u001b[0mz_k\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlogdet\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mflow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mz_0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     30\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     31\u001b[0m         \u001b[0mx_decoded\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdecode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mz_k\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/common/software/pytorch/1.8.0-gpu/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    888\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-36-9eb40dffb624>\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m     20\u001b[0m         \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     21\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayers\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 22\u001b[0;31m             \u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlogdet\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayers\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     23\u001b[0m             \u001b[0;31m# revert the dimension of the output after each block\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     24\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0muse_revert\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/common/software/pytorch/1.8.0-gpu/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    888\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/global/u2/a/agarabag/DarkFlow/layers.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    219\u001b[0m         \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    220\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mblock\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 221\u001b[0;31m             \u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlogdet\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mblock\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    222\u001b[0m             \u001b[0mlogdetSum\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mlogdet\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    223\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/common/software/pytorch/1.8.0-gpu/lib/python3.8/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m_call_impl\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    887\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    888\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 889\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    890\u001b[0m         for hook in itertools.chain(\n\u001b[1;32m    891\u001b[0m                 \u001b[0m_global_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/global/u2/a/agarabag/DarkFlow/layers.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m    189\u001b[0m                 \u001b[0mactivation\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mactivation\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munsqueeze\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    190\u001b[0m                 \u001b[0mactivation_gradient\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mactivation_gradient\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munsqueeze\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 191\u001b[0;31m             \u001b[0moutput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mactivation\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdiag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mscale\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    192\u001b[0m             \u001b[0mactivation_gradient\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mactivation_gradient\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdiag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mscale\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    193\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mRuntimeError\u001b[0m: Tensor for 'out' is on CPU, Tensor for argument #1 'self' is on CPU, but expected them to be on GPU (while checking arguments for addmm)"
     ]
    }
   ],
   "source": [
    "fake_SRs = []\n",
    "for i in range(len(list_of_cgen_models)):\n",
    "\n",
    "    if np.abs(cond_max) > 0:\n",
    "        train_mjj_vals_scaled = train_mjj_vals/cond_max\n",
    "    else:\n",
    "        train_mjj_vals_scaled = train_mjj_vals\n",
    "\n",
    "    print (\"TTTTTT: \", train_mjj_vals.shape[0], train_mjj_vals.shape, train_mjj_vals_scaled.shape)\n",
    "    fake_SR_t = generate_cond_data(i,train_mjj_vals.shape[0],train_mjj_vals_scaled)\n",
    "    # fake_SR = fake_SR_t.cpu().detach().numpy()\n",
    "\n",
    "    for i in range(0,fake_SR.shape[1]):\n",
    "    \tfake_SR[:,i]=fake_SR[:,i]*x_max[i]\n",
    "\n",
    "    fake_SRs.append(fake_SR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "fake_SRs = np.array(fake_SRs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 400000, 4)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fake_SRs.shape\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.save(\"fake_test_SR_cbvae_n400000.npy\", fake_SRs[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# cfake_sr = generate_cond_data(0,y_innerdata_train.shape[0],scaler_mjj.transform(y_innerdata_train.reshape(y_innerdata_train.shape[0],1)))\n",
    "fake_vs_real([fake_SRs[0]],[X_innerdata_train],0,[\"\"])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "fake_SBs = []\n",
    "for i in range(len(list_of_cgen_models)):\n",
    "\n",
    "    fake_SB = generate_cond_data(i,y_data.shape[0],y_data)\n",
    "\n",
    "    for i in range(0,fake_SB.shape[1]):\n",
    "    \tfake_SB[:,i]=fake_SB[:,i]*x_max[i]\n",
    "\n",
    "    fake_SBs.append(fake_SB)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "fake_SBs = np.array(fake_SBs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(878648, 4)\n"
     ]
    }
   ],
   "source": [
    "print(fake_SBs[0].shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fake_vs_real([fake_SBs[0]],[X_data_unscaled],0,[\"\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "fileNameNN = '/workdir/huichi/DeepEvents/LHCO2020/4var/LHCO2020_cB-VAE_events.csv'\n",
    "dfNN = pd.read_csv(fileNameNN, delimiter = ' ', header=None, index_col=False)\n",
    "dNN = dfNN.to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x720 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fake_vs_real([dNN],[X_data_unscaled],0,[\"\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'x_fake_SR' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_4279/2693133618.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mbkg_models\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mx_fake_SR\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx_real_SR\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0mbkg_names\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m\"GAN\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"MC\"\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'x_fake_SR' is not defined"
     ]
    }
   ],
   "source": [
    "# Loop over signal size\n",
    "aucs = []\n",
    "rocs = []\n",
    "losses = []\n",
    "val_losses = []\n",
    "accs = []\n",
    "val_accs = []\n",
    "num_epoch=1000\n",
    "batch_size=5096\n",
    "\n",
    "ts = 60000\n",
    "\n",
    "bkg_models = [x_fake_SR, x_real_SR]\n",
    "bkg_names = [\"GAN\", \"MC\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'bkg_models' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m/tmp/ipykernel_4279/4210146503.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;32mfor\u001b[0m \u001b[0mbackground\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mbkg_models\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m     \u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_test\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mprep_and_shufflesplit_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbackground\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbackground\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx_sig_SR\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msize_each\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mts\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mshuffle_seed\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m62\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0.2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtest_size_each\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m10000\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m     \u001b[0mdense_sizes\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m     \u001b[0mmodel\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mDNN\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_dim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m8\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdropouts\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ml2_regs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1e-4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdense_sizes\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdense_sizes\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msummary\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0mauc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mroc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0macc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mval_acc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtrainingNN\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_epoch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m100\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_val\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mY_test\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mNameError\u001b[0m: name 'bkg_models' is not defined"
     ]
    }
   ],
   "source": [
    "for background in bkg_models:\n",
    "    X_train, X_val, X_test, Y_train, Y_val, Y_test = prep_and_shufflesplit_data(background, background, x_sig_SR, 1, size_each = ts, shuffle_seed = 62, val = 0.2, test_size_each = 10000)\n",
    "    dense_sizes = (100,100)\n",
    "    model = DNN(input_dim=8, dropouts=0.2, l2_regs=1e-4, dense_sizes=dense_sizes, summary=False)\n",
    "    auc, roc, loss, acc, val_loss, val_acc = trainingNN(model, num_epoch, 100, batch_size, X_train, X_val, X_test, Y_train, Y_val, Y_test)\n",
    "    rocs.append(roc)\n",
    "    aucs.append(auc)\n",
    "    losses.append(loss)\n",
    "    accs.append(acc)\n",
    "    val_losses.append(val_loss)\n",
    "    val_accs.append(val_acc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
  },
  "kernelspec": {
   "display_name": "TOARCH",
   "language": "python",
   "name": "torch"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}