quickstart.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# A brief explanation of the revamped CBRAIN module\n",
    "\n",
    "In this guide, we will go through the steps required to preprocess the raw climate model output, train a neural network and then analyze how good it is."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cbrain.imports import *\n",
    "from cbrain.cam_constants import *\n",
    "from cbrain.utils import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing\n",
    "\n",
    "Preprocessing works using the `preprocessing.py` script in the main directory of the repository along with a configuration file. Let's have a look at one such configuration file to see what is required."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is `000_test.yml`:\n",
    "\n",
    "```yaml\n",
    "vars : [QBP, QCBP, QIBP, TBP, VBP, PS, SOLIN, SHFLX, LHFLX, PHQ, PHCLDLIQ, PHCLDICE, TPHYSTND, QRL, QRS, DTVKE, FSNT, FSNS, FLNT, FLNS, PRECT, PRECTEND, PRECST, PRECSTEN, Qdt_adiabatic, QCdt_adiabatic, QIdt_adiabatic, Tdt_adiabatic, Vdt_adiabatic]\n",
    "\n",
    "in_dir : /local/S.Rasp/sp32fbp_andkua/\n",
    "in_fns : AndKua_aqua_SPCAM3.0_sp_fbp32.cam2.h1.0000-01-0*-00000.nc\n",
    "\n",
    "out_dir: /local/S.Rasp/preprocessed_data/\n",
    "out_fn: 000_train.nc\n",
    "\n",
    "val_in_fns: AndKua_aqua_SPCAM3.0_sp_fbp32.cam2.h1.0000-02-0*-00000.nc\n",
    "val_out_fn: 000_valid.nc\n",
    "\n",
    "norm_fn: 000_norm.nc\n",
    "```\n",
    "\n",
    "One new thing is that it is not necessary to specifically state the input and output variables at the preprocessing script. These will be chosen later in the data generator. This mean that `vars` can contain as many variables as possible. This should reduce the number of times the preprocessing has to be run.\n",
    "\n",
    "Some of the variables are saved directly in the model output, e.g. `PHQ`, while others, e.g. `QBP` have to be derived. Currently, the following derived variables are available: `*dt_adiabatic`, `*BP` and `PRECST`. If you want to add others, you have to implement it in `convert_dataset.py`.\n",
    "\n",
    "`in_dir` and `in_fns` describe the raw climate model file names. `out_dir` and `out_fn` denote the path to the preprocessed file. \n",
    "\n",
    "The training dataset will additionally be preshuffled.\n",
    "\n",
    "If `val_*` is given, a separate validation dataset will be created.\n",
    "\n",
    "If `norm_fn` is given, statistics will be computed from the training dataset, e.g. the mean and standard deviation.  Note that for large training files this takes a very, very long time. Therefore, I recommend computing the normalization file on a small dataset and use it for the large training dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2019-04-03 14:52:26.858479: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA\n",
      "2019-04-03 14:52:27.047180: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Found device 0 with properties: \n",
      "name: GeForce GTX 1080 major: 6 minor: 1 memoryClockRate(GHz): 1.7335\n",
      "pciBusID: 0000:b3:00.0\n",
      "totalMemory: 7.93GiB freeMemory: 6.04GiB\n",
      "2019-04-03 14:52:27.047219: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1511] Adding visible gpu devices: 0\n",
      "2019-04-03 14:52:39.252462: I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
      "2019-04-03 14:52:39.253048: I tensorflow/core/common_runtime/gpu/gpu_device.cc:988]      0 \n",
      "2019-04-03 14:52:39.253084: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1001] 0:   N \n",
      "2019-04-03 14:52:39.253949: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 5814 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080, pci bus id: 0000:b3:00.0, compute capability: 6.1)\n",
      "04/03/2019 02:52:39 PM Preprocess training dataset\n",
      "04/03/2019 02:52:39 PM Start preprocessing file /local/S.Rasp/preprocessed_data/000_train.nc\n",
      "04/03/2019 02:52:39 PM Reading input files\n",
      "04/03/2019 02:52:40 PM Crop levels\n",
      "04/03/2019 02:52:40 PM Create stacked dataarray\n",
      "These time steps are cut: []\n",
      "04/03/2019 02:52:40 PM Stack and reshape dataarray\n",
      "04/03/2019 02:52:43 PM Save dataarray as /local/S.Rasp/preprocessed_data/000_train.nc\n",
      "04/03/2019 02:57:17 PM Done!\n",
      "04/03/2019 02:57:17 PM Shuffle training dataset\n",
      "04/03/2019 02:57:17 PM Start shuffling /local/S.Rasp/preprocessed_data/000_train.nc into /local/S.Rasp/preprocessed_data/000_train_shuffle.nc. Open and create datasets.\n",
      "04/03/2019 02:57:18 PM Shuffle!\n",
      "100%|█████████████████████████████████████████████| 1/1 [00:41<00:00, 41.28s/it]\n",
      "04/03/2019 02:57:59 PM Closing datasets\n",
      "04/03/2019 02:57:59 PM Done!\n",
      "04/03/2019 02:57:59 PM Preprocess validation dataset\n",
      "04/03/2019 02:57:59 PM Start preprocessing file /local/S.Rasp/preprocessed_data/000_valid.nc\n",
      "04/03/2019 02:57:59 PM Reading input files\n",
      "04/03/2019 02:58:06 PM Crop levels\n",
      "04/03/2019 02:58:06 PM Create stacked dataarray\n",
      "These time steps are cut: []\n",
      "04/03/2019 02:58:07 PM Stack and reshape dataarray\n",
      "04/03/2019 02:58:10 PM Save dataarray as /local/S.Rasp/preprocessed_data/000_valid.nc\n",
      "04/03/2019 03:02:38 PM Done!\n",
      "04/03/2019 03:02:38 PM Compute normalization file from train\n",
      "04/03/2019 03:02:38 PM Start normalizing by creating dataset.\n",
      "04/03/2019 03:02:38 PM Compute standard normalizations.\n",
      "04/03/2019 03:05:30 PM Saving normalization file as /local/S.Rasp/preprocessed_data/000_norm.nc\n",
      "04/03/2019 03:05:30 PM Done!\n",
      "04/03/2019 03:05:31 PM Finish entire preprocessing script.\n"
     ]
    }
   ],
   "source": [
    "!python preprocessing.py -c pp_config/000_test.yml"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/local/S.Rasp/preprocessed_data/000_norm.nc\r\n",
      "/local/S.Rasp/preprocessed_data/000_train.nc\r\n",
      "/local/S.Rasp/preprocessed_data/000_train_shuffle.nc\r\n",
      "/local/S.Rasp/preprocessed_data/000_valid.nc\r\n"
     ]
    }
   ],
   "source": [
    "!ls /local/S.Rasp/preprocessed_data/000*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## Training the model step-by-step\n",
    "\n",
    "Let's now go through the steps to train a neural network starting with the new data generator."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Data generator\n",
    "\n",
    "To read the preprocessed file and feed the data to the neural net, we will use the `DataGenerator` class in the cbrain module. \n",
    "\n",
    "At this stage we will not define the variables we want in the input and output of the neural network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "in_vars = ['QBP', 'QCBP', 'QIBP', 'TBP', 'VBP', \n",
    "           'Qdt_adiabatic', 'QCdt_adiabatic', 'QIdt_adiabatic', 'Tdt_adiabatic', 'Vdt_adiabatic',\n",
    "           'PS', 'SOLIN', 'SHFLX', 'LHFLX']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "out_vars = ['PHQ', 'PHCLDLIQ', 'PHCLDICE', 'TPHYSTND', 'QRL', 'QRS', 'DTVKE', \n",
    "            'FSNT', 'FSNS', 'FLNT', 'FLNS', 'PRECT', 'PRECTEND', 'PRECST', 'PRECSTEN']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Next we need to think about how we want to normalize/scale the inputs and outputs. \n",
    "\n",
    "For the inputs we will use a pretty standard normalization, which we will get to later.\n",
    "\n",
    "As for the outputs, we would like to scale each variable physically. For this we will create a dictionary, which will contain a factor or a vector of factors for each variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "scale_dict = {\n",
    "    'PHQ': L_V/G, \n",
    "    'PHCLDLIQ': L_V/G, \n",
    "    'PHCLDICE': L_V/G, \n",
    "    'TPHYSTND': C_P/G, \n",
    "    'QRL': C_P/G, \n",
    "    'QRS': C_P/G, \n",
    "    'DTVKE': C_P/G, \n",
    "    'FSNT': 1, \n",
    "    'FSNS': 1, \n",
    "    'FLNT': 1, \n",
    "    'FLNS': 1, \n",
    "    'PRECT': RHO_L*L_V, \n",
    "    'PRECTEND': 1e-3*RHO_L*L_V, \n",
    "    'PRECST': RHO_L*L_V, \n",
    "    'PRECSTEN': 1e-3*RHO_L*L_V\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Takes representative value for PS since purpose is normalization\n",
    "PS = 1e5; P0 = 1e5;\n",
    "P = P0*hyai+PS*hybi; # Total pressure [Pa]\n",
    "dP = P[1:]-P[:-1]; # Differential pressure [Pa]  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "for v in ['PHQ', 'PHCLDLIQ', 'PHCLDICE', 'TPHYSTND', 'QRL', 'QRS']:\n",
    "    scale_dict[v] *= dP\n",
    "scale_dict['DTVKE'] *= (dP/DT)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Now we will save this dictionary as a pickle file, so that we can later load it in th training script. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "save_pickle('./nn_config/scale_dicts/001_toms_scaling.pkl', scale_dict)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "With this dictionary, we can now use create a DataGenerator instance.\n",
    "\n",
    "A word about the normalization. The normalization is handled by Normalizer classes, defined in `normalization.py`. As default in `DataGenerator`, for input normalization the `InputNormalizer` class is used, while the outputs are scaled using the `DictNormalizer`. `DataGenerator` takes a tuple of strings for `input_transform`. This tuple describes which arrays from the normalization file (`norm_fn`) the input will be subtracted and divided by. \n",
    "\n",
    "If you want to create your own fancy normalization, you have to create such a Normalizer which has to have a `transform` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "from cbrain.data_generator import DataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "train_gen = DataGenerator(\n",
    "    data_fn = '/local/S.Rasp/preprocessed_data/000_train_shuffle.nc',\n",
    "    input_vars = in_vars,\n",
    "    output_vars = out_vars,\n",
    "    norm_fn = '/local/S.Rasp/preprocessed_data/000_norm.nc',\n",
    "    input_transform = ('mean', 'maxrs'),\n",
    "    output_transform = scale_dict,\n",
    "    batch_size=1024,\n",
    "    shuffle=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1024, 304), (1024, 218))"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X, Y = train_gen[0]; X.shape, Y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Create a model with conservation layers\n",
    "\n",
    "Next we need to create a model. This is just basic Keras. I will show here how to use the conservation layers written by Tom. These layers require some additional input, in particular the normalization information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "from cbrain.layers import *\n",
    "from tensorflow.keras.layers import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "inp_layer = Input(shape=(304,))\n",
    "x = Dense(214, activation='elu')(inp_layer)\n",
    "x = SurRadLayer(\n",
    "    inp_div=train_gen.input_transform.div,\n",
    "    inp_sub=train_gen.input_transform.sub,\n",
    "    norm_q=scale_dict['PHQ']\n",
    ")([inp_layer, x])\n",
    "x = MassConsLayer(\n",
    "    inp_div=train_gen.input_transform.div,\n",
    "    inp_sub=train_gen.input_transform.sub,\n",
    "    norm_q=scale_dict['PHQ']\n",
    ")([inp_layer, x])\n",
    "out_layer = EntConsLayer(\n",
    "    inp_div=train_gen.input_transform.div,\n",
    "    inp_sub=train_gen.input_transform.sub,\n",
    "    norm_q=scale_dict['PHQ']\n",
    ")([inp_layer, x])\n",
    "model = tf.keras.models.Model(inp_layer, out_layer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 304)          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 214)          65270       input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "sur_rad_layer (SurRadLayer)     (None, 216)          0           input_1[0][0]                    \n",
      "                                                                 dense[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "mass_cons_layer (MassConsLayer) (None, 217)          0           input_1[0][0]                    \n",
      "                                                                 sur_rad_layer[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "ent_cons_layer (EntConsLayer)   (None, 218)          0           input_1[0][0]                    \n",
      "                                                                 mass_cons_layer[0][0]            \n",
      "==================================================================================================\n",
      "Total params: 65,270\n",
      "Trainable params: 65,270\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "### Train the model with the weakly constrained energy loss\n",
    "\n",
    "Again, this is basic Keras. Tom implemented the option to penalize the network for violating physical constraints, however, which we will do here.\n",
    "\n",
    "I implemented the weakly constrained loss as a class in `losses.py`. When initializing this class, you need to pass the input_tensor from the model and again the normalization information.\n",
    "\n",
    "This loss computes three losses internally: the standard MSE, a loss for mass conservation and a loss for enthalpy conservation. You can weight the losses by changing the `alpha_*` parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "from cbrain.losses import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "weak_loss = WeakLoss(inp_layer, inp_div=train_gen.input_transform.div, inp_sub=train_gen.input_transform.sub,\n",
    "                     norm_q=scale_dict['PHQ'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "mass_loss = WeakLoss(inp_layer, inp_div=train_gen.input_transform.div, inp_sub=train_gen.input_transform.sub,\n",
    "                     norm_q=scale_dict['PHQ'], alpha_mass=1, alpha_ent=0, name='mass_loss')\n",
    "ent_loss = WeakLoss(inp_layer, inp_div=train_gen.input_transform.div, inp_sub=train_gen.input_transform.sub,\n",
    "                     norm_q=scale_dict['PHQ'], alpha_mass=0, alpha_ent=1, name='ent_loss')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "model.compile(tf.keras.optimizers.Adam(lr=0.01), loss=weak_loss, metrics=[mass_loss, ent_loss, mse])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "hidden": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "3448/3448 [==============================] - 109s 32ms/step - loss: 740.1542 - mass_loss: 3.6446e-11 - ent_loss: 7.8155e-11 - mean_squared_error: 1480.3085\n",
      "Epoch 2/5\n",
      "3448/3448 [==============================] - 67s 20ms/step - loss: 540.9411 - mass_loss: 6.3662e-11 - ent_loss: 1.4316e-10 - mean_squared_error: 1081.8822\n",
      "Epoch 3/5\n",
      "3448/3448 [==============================] - 67s 19ms/step - loss: 419.6522 - mass_loss: 8.7082e-11 - ent_loss: 1.6894e-10 - mean_squared_error: 839.3044\n",
      "Epoch 4/5\n",
      "3448/3448 [==============================] - 66s 19ms/step - loss: 356.3085 - mass_loss: 1.0223e-10 - ent_loss: 1.9572e-10 - mean_squared_error: 712.6170\n",
      "Epoch 5/5\n",
      "3448/3448 [==============================] - 67s 19ms/step - loss: 322.0714 - mass_loss: 1.1291e-10 - ent_loss: 2.1134e-10 - mean_squared_error: 644.1428\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7fd0fc0740b8>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit_generator(train_gen, epochs=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "hidden": true
   },
   "source": [
    "Naturally, since we are using a conserving network, the conservation losses are basically zero. Puh..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the network using the train.py script\n",
    "\n",
    "Doing the training in a notebook is good for experimentation but for testing different configurations and using large training datasets, we need a command line script, which is called `train.py`. So, let's check out how to use it.\n",
    "\n",
    "Again we will create a configuration file."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is `000_example.yml`:\n",
    "\n",
    "```\n",
    "# Example training configuration file\n",
    "exp_name: 000_example\n",
    "data_dir: /local/S.Rasp/preprocessed_data/\n",
    "train_fn: 000_train_shuffle.nc\n",
    "valid_fn: 000_valid.nc\n",
    "norm_fn: 000_norm.nc\n",
    "\n",
    "inputs: [QBP, QCBP, QIBP, TBP, VBP, Qdt_adiabatic, QCdt_adiabatic, QIdt_adiabatic, Tdt_adiabatic, Vdt_adiabatic, PS, SOLIN, SHFLX, LHFLX]\n",
    "outputs: [PHQ, PHCLDLIQ, PHCLDICE, TPHYSTND, QRL, QRS, DTVKE, FSNT, FSNS, FLNT, FLNS, PRECT, PRECTEND, PRECST, PRECSTEN]\n",
    "\n",
    "input_sub: mean\n",
    "input_div: maxrs\n",
    "output_dict: /home/s/S.Rasp/repositories/CBRAIN-CAM/nn_config/scale_dicts/001_toms_scaling.pkl\n",
    "\n",
    "hidden_layers: [128, 214]\n",
    "epochs: 10\n",
    "conservation_layer: True\n",
    "loss: weak_loss\n",
    "```\n",
    "\n",
    "Most of the arguments are pretty self-explanatory (I hope). If you are confused look at the definition at the bottom of `train.py`.\n",
    "\n",
    "This script automatically uses learning rate decay.\n",
    "\n",
    "Additionally, and very importantly for the eventual implementation in CAM, it also saves the network and weights."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2019-04-03 15:23:20.846557: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA\n",
      "2019-04-03 15:23:20.982342: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1432] Found device 0 with properties: \n",
      "name: GeForce GTX 1080 major: 6 minor: 1 memoryClockRate(GHz): 1.7335\n",
      "pciBusID: 0000:b3:00.0\n",
      "totalMemory: 7.93GiB freeMemory: 5.87GiB\n",
      "2019-04-03 15:23:20.982378: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1511] Adding visible gpu devices: 0\n",
      "2019-04-03 15:23:21.453378: I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
      "2019-04-03 15:23:21.453417: I tensorflow/core/common_runtime/gpu/gpu_device.cc:988]      0 \n",
      "2019-04-03 15:23:21.453425: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1001] 0:   N \n",
      "2019-04-03 15:23:21.453721: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 5642 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080, pci bus id: 0000:b3:00.0, compute capability: 6.1)\n",
      "04/03/2019 03:23:21 PM Create training and validation data generators\n",
      "04/03/2019 03:23:21 PM Build model\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            (None, 304)          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "dense (Dense)                   (None, 256)          78080       input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "leaky_re_lu (LeakyReLU)         (None, 256)          0           dense[0][0]                      \n",
      "__________________________________________________________________________________________________\n",
      "dense_1 (Dense)                 (None, 256)          65792       leaky_re_lu[0][0]                \n",
      "__________________________________________________________________________________________________\n",
      "leaky_re_lu_1 (LeakyReLU)       (None, 256)          0           dense_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 256)          65792       leaky_re_lu_1[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "leaky_re_lu_2 (LeakyReLU)       (None, 256)          0           dense_2[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_3 (Dense)                 (None, 214)          54998       leaky_re_lu_2[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "leaky_re_lu_3 (LeakyReLU)       (None, 214)          0           dense_3[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "sur_rad_layer (SurRadLayer)     (None, 216)          0           input_1[0][0]                    \n",
      "                                                                 leaky_re_lu_3[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "mass_cons_layer (MassConsLayer) (None, 217)          0           input_1[0][0]                    \n",
      "                                                                 sur_rad_layer[0][0]              \n",
      "__________________________________________________________________________________________________\n",
      "ent_cons_layer (EntConsLayer)   (None, 218)          0           input_1[0][0]                    \n",
      "                                                                 mass_cons_layer[0][0]            \n",
      "==================================================================================================\n",
      "Total params: 264,662\n",
      "Trainable params: 264,662\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n",
      "None\n",
      "04/03/2019 03:23:21 PM Compile model\n",
      "04/03/2019 03:23:22 PM Train model\n",
      "2019-04-03 15:23:23.590239: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1511] Adding visible gpu devices: 0\n",
      "2019-04-03 15:23:23.590287: I tensorflow/core/common_runtime/gpu/gpu_device.cc:982] Device interconnect StreamExecutor with strength 1 edge matrix:\n",
      "2019-04-03 15:23:23.590296: I tensorflow/core/common_runtime/gpu/gpu_device.cc:988]      0 \n",
      "2019-04-03 15:23:23.590305: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1001] 0:   N \n",
      "2019-04-03 15:23:23.590595: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 5642 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1080, pci bus id: 0000:b3:00.0, compute capability: 6.1)\n",
      "Learning rate = 0.01\n",
      "Epoch 1/10\n",
      "3448/3448 [==============================] - 89s 26ms/step - loss: 66.6859 - mean_squared_error: 133.3718 - mass_loss: 3.4461e-10 - ent_loss: 3.2758e-10 - val_loss: 51.0592 - val_mean_squared_error: 102.1183 - val_mass_loss: 3.8388e-10 - val_ent_loss: 1.9052e-10\n",
      "Learning rate = 0.01\n",
      "Epoch 2/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 50.0245 - mean_squared_error: 100.0490 - mass_loss: 3.7326e-10 - ent_loss: 3.3711e-10 - val_loss: 50.6125 - val_mean_squared_error: 101.2250 - val_mass_loss: 4.0006e-10 - val_ent_loss: 2.0463e-10\n",
      "Learning rate = 0.002\n",
      "Epoch 3/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 44.0968 - mean_squared_error: 88.1936 - mass_loss: 3.8555e-10 - ent_loss: 3.3528e-10 - val_loss: 45.2651 - val_mean_squared_error: 90.5302 - val_mass_loss: 3.9009e-10 - val_ent_loss: 1.8705e-10\n",
      "Learning rate = 0.002\n",
      "Epoch 4/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 43.1805 - mean_squared_error: 86.3610 - mass_loss: 3.9110e-10 - ent_loss: 3.3402e-10 - val_loss: 44.7589 - val_mean_squared_error: 89.5178 - val_mass_loss: 3.6946e-10 - val_ent_loss: 1.7980e-10\n",
      "Learning rate = 0.0004000000000000001\n",
      "Epoch 5/10\n",
      "3448/3448 [==============================] - 83s 24ms/step - loss: 41.5774 - mean_squared_error: 83.1547 - mass_loss: 3.9188e-10 - ent_loss: 3.3378e-10 - val_loss: 43.3000 - val_mean_squared_error: 86.5999 - val_mass_loss: 4.0520e-10 - val_ent_loss: 1.9103e-10\n",
      "Learning rate = 0.0004000000000000001\n",
      "Epoch 6/10\n",
      "3448/3448 [==============================] - 83s 24ms/step - loss: 41.2948 - mean_squared_error: 82.5897 - mass_loss: 3.9295e-10 - ent_loss: 3.3513e-10 - val_loss: 43.2729 - val_mean_squared_error: 86.5458 - val_mass_loss: 3.8670e-10 - val_ent_loss: 1.8113e-10\n",
      "Learning rate = 8.000000000000002e-05\n",
      "Epoch 7/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 40.8886 - mean_squared_error: 81.7773 - mass_loss: 3.9385e-10 - ent_loss: 3.3514e-10 - val_loss: 42.9063 - val_mean_squared_error: 85.8126 - val_mass_loss: 4.0482e-10 - val_ent_loss: 1.8694e-10\n",
      "Learning rate = 8.000000000000002e-05\n",
      "Epoch 8/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 40.8127 - mean_squared_error: 81.6254 - mass_loss: 3.9597e-10 - ent_loss: 3.3652e-10 - val_loss: 42.8711 - val_mean_squared_error: 85.7422 - val_mass_loss: 4.0361e-10 - val_ent_loss: 1.8883e-10\n",
      "Learning rate = 1.6000000000000003e-05\n",
      "Epoch 9/10\n",
      "3448/3448 [==============================] - 83s 24ms/step - loss: 40.7166 - mean_squared_error: 81.4333 - mass_loss: 3.9348e-10 - ent_loss: 3.3688e-10 - val_loss: 42.8298 - val_mean_squared_error: 85.6596 - val_mass_loss: 4.0140e-10 - val_ent_loss: 1.8511e-10\n",
      "Learning rate = 1.6000000000000003e-05\n",
      "Epoch 10/10\n",
      "3448/3448 [==============================] - 82s 24ms/step - loss: 40.7036 - mean_squared_error: 81.4072 - mass_loss: 3.9575e-10 - ent_loss: 3.3639e-10 - val_loss: 42.8211 - val_mean_squared_error: 85.6422 - val_mass_loss: 4.0147e-10 - val_ent_loss: 1.8437e-10\n",
      "04/03/2019 03:37:14 PM Saving model as ./saved_models/000_example/model.h5\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "04/03/2019 03:37:15 PM Saving weights as ./saved_models/000_example/weights.h5\n",
      "04/03/2019 03:37:15 PM Done!\n"
     ]
    }
   ],
   "source": [
    "!python train.py -c nn_config/000_example.yml"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see that we are overfitting quite a bit which is to be expected with such a small dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "inp_div.txt\t   layer2_bias.txt    layer4_bias.txt\t weights.h5\r\n",
      "inp_sub.txt\t   layer2_kernel.txt  layer4_kernel.txt\r\n",
      "layer1_bias.txt    layer3_bias.txt    model.h5\r\n",
      "layer1_kernel.txt  layer3_kernel.txt  out_scale.txt\r\n"
     ]
    }
   ],
   "source": [
    "!ls ./saved_models/000_example/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model diagnostics\n",
    "\n",
    "Finally, we would like to know how well our model does in more detail that just looking at the loss.\n",
    "\n",
    "For this I wrote the `ModelDiagnostics` class. It is designed to be convenient. Let's see what it can do.\n",
    "\n",
    "For basic usage it only needs two arguments: First, the configuration file used for neural network training and second the data file which is to be used for validation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cbrain.model_diagnostics import ModelDiagnostics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "md = ModelDiagnostics('nn_config/000_example.yml', '/local/S.Rasp/preprocessed_data/000_valid.nc')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting\n",
    "\n",
    "The first thing we can do is plot the truth alongside the model prediction. These functions take all the regular matplotlib arguments."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "md.plot_double_yz(itime=0, ilon=0, var='PHQ', vmin=-8e-8, vmax=8e-8, cmap='bwr');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x360 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "md.plot_double_xy(0, 0, 'PRECT');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Compute statistics\n",
    "\n",
    "Last but not least, we can compute statistics over the entire validation dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fd8e2b84970e41af9757cb0ace9d8f99",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=431), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/s/S.Rasp/repositories/CBRAIN-CAM/cbrain/model_diagnostics/model_diagnostics.py:130: RuntimeWarning: divide by zero encountered in true_divide\n",
      "  self.stats['r2'] = 1. - (self.stats['mse'] / self.stats['true_var'])\n",
      "/home/s/S.Rasp/repositories/CBRAIN-CAM/cbrain/model_diagnostics/model_diagnostics.py:136: RuntimeWarning: divide by zero encountered in true_divide\n",
      "  self.stats['hor_r2'] = 1 - (self.stats['hor_mse'] / self.stats['hor_tvar'])\n"
     ]
    }
   ],
   "source": [
    "md.compute_stats()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now there is a stats dictionary containing many statistics. If you want to implement your own statistics, you wil have to do so in the `compute_stats` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['bias', 'mse', 'pred_mean', 'true_mean', 'pred_sqmean', 'true_sqmean', 'pred_var', 'true_var', 'r2', 'hor_tsqmean', 'hor_tmean', 'hor_mse', 'hor_tvar', 'hor_r2'])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "md.stats.keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "One common statistic, for example, is the R2. So let's plot that. The R2 is averaged over time but all other dimensions are still available. Further, the vertical level is still stacked, so we will have to figure out the variable indices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64, 128, 218)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "md.stats['r2'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x7fd038130908>"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 614.4x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(md.stats['r2'][:, :, md.get_output_var_idx('PHQ')].mean(1).T, vmin=0, vmax=1, cmap='Spectral')\n",
    "plt.colorbar(shrink=0.7)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also like looking at the horizontally averaged R2 for each variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(218,)"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "md.stats['hor_r2'].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['PHQ', 'PHCLDLIQ', 'PHCLDICE', 'TPHYSTND', 'QRL', 'QRS', 'DTVKE']"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "md.valid_gen.output_vars[:7]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the vertical coordinate in pressure levels\n",
    "PS = 1e5; P0 = 1e5;\n",
    "P = P0*hyai+PS*hybi; # Total pressure [Pa]\n",
    "P = (P[1:] + P[:-1]) / 2 / 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fcff77317f0>"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(8, 8))\n",
    "for v in md.valid_gen.output_vars[:7]:\n",
    "    ax.plot(md.stats['hor_r2'][md.get_output_var_idx(v)], P, label=v)\n",
    "ax.set_xlim(0, 1)\n",
    "ax.invert_yaxis()\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Last words\n",
    "\n",
    "So this is the new version of cbrain. Feel free to fork you own branch and make changes. If you think this might be useful for others, just create a pull request.\n",
    "\n",
    "Let me know if you have questions,\n",
    "Stephan"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": false,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}