{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DeepLDA: Alanine dipeptide and aldol reaction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Reference paper: _Bonati, Rizzi and Parrinello, [JCPL](https://pubs.acs.org/doi/10.1021/acs.jpclett.0c00535) (2020)_ [[arXiv]](https://arxiv.org/abs/2002.06562).\n",
    "\n",
    "Prerequisite: DeepLDA tutorial.\n",
    "\n",
    "[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/luigibonati/mlcolvar/blob/main/docs/notebooks/examples/ex_DeepLDA.ipynb)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<torch._C.Generator at 0x7f5ae82043b0>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Colab setup\n",
    "import os\n",
    "\n",
    "if os.getenv(\"COLAB_RELEASE_TAG\"):\n",
    "    import subprocess\n",
    "    subprocess.run('wget https://raw.githubusercontent.com/luigibonati/mlcolvar/main/colab_setup.sh', shell=True)\n",
    "    cmd = subprocess.run('bash colab_setup.sh EXAMPLE', shell=True, stdout=subprocess.PIPE)\n",
    "    print(cmd.stdout.decode('utf-8'))\n",
    "\n",
    "# IMPORT PACKAGES\n",
    "import torch\n",
    "import lightning\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Set seed for reproducibility\n",
    "torch.manual_seed(41)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Alanine dipeptide"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We first train a DeepLDA CV on the alanine dipeptide data, which is one of the examples of the [DeepLDA paper](https://pubs.acs.org/doi/10.1021/acs.jpclett.0c00535). This is a small molecule often used as a benchmark for enhanced sampling, as it has two metastable states which are well described by the Ramachandran angles $\\phi$ and $\\psi$. To build a general mode, we will not use this information and just use as input features the 45 distances between heavy atoms.\n",
    "\n",
    "<center><img src=\"images/alanine.png\" width=\"500\" /></center>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We use the alanine dipeptide simulation data from the [PLUMED-MASTERCLASS](https://github.com/luigibonati/masterclass-plumed/) repository."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class 0 dataframe shape:  (5001, 53)\n",
      "Class 1 dataframe shape:  (5001, 53)\n",
      "\n",
      " - Loaded dataframe (10002, 53): ['time', 'phi', 'psi', 'theta', 'xi', 'ene', 'd_2_5', 'd_2_6', 'd_2_7', 'd_2_9', 'd_2_11', 'd_2_15', 'd_2_16', 'd_2_17', 'd_2_19', 'd_5_6', 'd_5_7', 'd_5_9', 'd_5_11', 'd_5_15', 'd_5_16', 'd_5_17', 'd_5_19', 'd_6_7', 'd_6_9', 'd_6_11', 'd_6_15', 'd_6_16', 'd_6_17', 'd_6_19', 'd_7_9', 'd_7_11', 'd_7_15', 'd_7_16', 'd_7_17', 'd_7_19', 'd_9_11', 'd_9_15', 'd_9_16', 'd_9_17', 'd_9_19', 'd_11_15', 'd_11_16', 'd_11_17', 'd_11_19', 'd_15_16', 'd_15_17', 'd_15_19', 'd_16_17', 'd_16_19', 'd_17_19', 'walker', 'labels']\n",
      " - Descriptors (10002, 45): ['d_2_5', 'd_2_6', 'd_2_7', 'd_2_9', 'd_2_11', 'd_2_15', 'd_2_16', 'd_2_17', 'd_2_19', 'd_5_6', 'd_5_7', 'd_5_9', 'd_5_11', 'd_5_15', 'd_5_16', 'd_5_17', 'd_5_19', 'd_6_7', 'd_6_9', 'd_6_11', 'd_6_15', 'd_6_16', 'd_6_17', 'd_6_19', 'd_7_9', 'd_7_11', 'd_7_15', 'd_7_16', 'd_7_17', 'd_7_19', 'd_9_11', 'd_9_15', 'd_9_16', 'd_9_17', 'd_9_19', 'd_11_15', 'd_11_16', 'd_11_17', 'd_11_19', 'd_15_16', 'd_15_17', 'd_15_19', 'd_16_17', 'd_16_19', 'd_17_19']\n"
     ]
    }
   ],
   "source": [
    "from mlcolvar.io import create_dataset_from_files\n",
    "from mlcolvar.data import DictModule\n",
    "\n",
    "filenames = [ \"https://raw.githubusercontent.com/luigibonati/masterclass-plumed/main/1_DeepLDA/0_unbiased-sA/COLVAR\", \n",
    "              \"https://raw.githubusercontent.com/luigibonati/masterclass-plumed/main/1_DeepLDA/0_unbiased-sB/COLVAR\" ]\n",
    "\n",
    "n_states = len(filenames)\n",
    "\n",
    "# load dataset\n",
    "dataset, df = create_dataset_from_files(filenames,\n",
    "                                        filter_args={'regex':'d_' }, # select distances between heavy atoms\n",
    "                                        create_labels=True,\n",
    "                                        return_dataframe=True, \n",
    "                                        )\n",
    "\n",
    "datamodule = DictModule(dataset,lengths=[0.8,0.2])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can inspect the distances through their histograms, from which we notice that no one is able to discriminate between the states."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x1000 with 45 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "descriptors_names = df.filter(regex='d_').columns.values\n",
    "\n",
    "fig,axs = plt.subplots(5,9,figsize=(20,10),sharey=True)\n",
    "\n",
    "for ax,desc in zip(axs.flatten(),descriptors_names):\n",
    "    df.pivot(columns='labels')[desc].plot.hist(bins=50,alpha=0.5,ax=ax,legend=False)\n",
    "    ax.set_title(desc)\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train DeepLDA "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DeepLDA(\n",
       "  (loss_fn): ReduceEigenvaluesLoss()\n",
       "  (norm_in): Normalization(in_features=45, out_features=45, mode=mean_std)\n",
       "  (nn): FeedForward(\n",
       "    (nn): Sequential(\n",
       "      (0): Linear(in_features=45, out_features=30, bias=True)\n",
       "      (1): ReLU(inplace=True)\n",
       "      (2): Linear(in_features=30, out_features=30, bias=True)\n",
       "      (3): ReLU(inplace=True)\n",
       "      (4): Linear(in_features=30, out_features=2, bias=True)\n",
       "    )\n",
       "  )\n",
       "  (lda): LDA(in_features=2, out_features=1)\n",
       ")"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from mlcolvar.cvs import DeepLDA\n",
    "\n",
    "n_input           = dataset['data'].shape[-1]\n",
    "nodes             = [n_input,30,30,2]\n",
    "# MODEL\n",
    "model = DeepLDA(nodes, n_states=n_states)\n",
    "\n",
    "model"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define trainer and fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True (cuda), used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/lightning/pytorch/trainer/connectors/logger_connector/logger_connector.py:67: UserWarning: Starting from v1.9.0, `tensorboardX` has been removed as a dependency of the `lightning.pytorch` package, due to potential conflicts with other packages in the ML ecosystem. For this reason, `logger=True` will use `CSVLogger` as the default logger, unless the `tensorboard` or `tensorboardX` packages are found. Please `pip install lightning[extra]` or one of them to enable TensorBoard support by default\n",
      "  warning_cache.warn(\n",
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/lightning/pytorch/loops/utilities.py:70: PossibleUserWarning: `max_epochs` was not set. Setting it to 1000 epochs. To train without an epoch limit, set `max_epochs=-1`.\n",
      "  rank_zero_warn(\n",
      "You are using a CUDA device ('NVIDIA GeForce RTX 3090') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision\n",
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
      "\n",
      "  | Name    | Type                  | Params | In sizes | Out sizes\n",
      "-------------------------------------------------------------------------\n",
      "0 | loss_fn | ReduceEigenvaluesLoss | 0      | ?        | ?        \n",
      "1 | norm_in | Normalization         | 0      | [45]     | [45]     \n",
      "2 | nn      | FeedForward           | 2.4 K  | [45]     | [2]      \n",
      "3 | lda     | LDA                   | 0      | [2]      | [1]      \n",
      "-------------------------------------------------------------------------\n",
      "2.4 K     Trainable params\n",
      "0         Non-trainable params\n",
      "2.4 K     Total params\n",
      "0.009     Total estimated model params size (MB)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                                           "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py:280: PossibleUserWarning: The number of training batches (1) is smaller than the logging interval Trainer(log_every_n_steps=50). Set a lower value for log_every_n_steps if you want to see logs for the training epoch.\n",
      "  rank_zero_warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 305: 100%|██████████| 1/1 [00:00<00:00, 17.84it/s, v_num=72] \n"
     ]
    }
   ],
   "source": [
    "from lightning.pytorch.callbacks.early_stopping import EarlyStopping\n",
    "from mlcolvar.utils.trainer import MetricsCallback\n",
    "\n",
    "# define callbacks\n",
    "metrics = MetricsCallback()\n",
    "early_stopping = EarlyStopping(monitor=\"valid_loss\", min_delta=0.1, patience=50)\n",
    "\n",
    "# define trainer\n",
    "trainer = lightning.Trainer(callbacks=[metrics, early_stopping],\n",
    "                     max_epochs=None, logger=None, enable_checkpointing=False)\n",
    "\n",
    "# fit\n",
    "trainer.fit( model, datamodule )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mlcolvar.utils.plot import plot_metrics\n",
    "\n",
    "ax = plot_metrics(metrics.metrics, \n",
    "                  keys=[x for x in  metrics.metrics.keys() if 'eigval' in x and 'epoch' in x],#['train_loss_epoch','valid_loss'],\n",
    "                  yscale='linear')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Analysis of the CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# evaluate cv on training set and append to dataframe\n",
    "X = dataset[:]['data']\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    s = model(torch.Tensor(X)).numpy()\n",
    "\n",
    "n_components = n_states-1\n",
    "\n",
    "for i in range(n_components):\n",
    "    df[f'CV{i}'] = s[:,i] "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As in the previous example, we can evaluate the histogram of the CV on the training set, from which we see how the two states are now discriminated. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,axs = plt.subplots(1,n_components,figsize=(5*n_components,4))\n",
    "if n_components == 1:\n",
    "    axs = [axs]\n",
    "\n",
    "for i,ax in enumerate(axs):\n",
    "    ax.hist(s[:,i],bins=100)\n",
    "    ax.set_xlabel(f'CV {i}')\n",
    "    ax.set_ylabel('Histogram')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Moreover, we can also visualize the CVs values on a physical space, e.g. in the Ramachandran plot ($\\phi$ and $\\psi$ space)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_components = n_states-1\n",
    "\n",
    "fig,axs = plt.subplots(1,n_components,figsize=(5*n_components,4))\n",
    "if n_components == 1:\n",
    "    axs = [axs]\n",
    "\n",
    "for i,ax in enumerate(axs):\n",
    "    df.plot.hexbin('phi','psi',C=f'CV{i}',cmap='fessa',ax=ax)\n",
    "    ax.set_title(f'CV {i} in phi/psi space')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By performing an enhanced sampling simulation along the CV we can we can compute the free energy profile and compare the isolines in the whole Ramachandran plot, which show how the DeepLDA CV captures approximately the minimum free energy pathway connecting the two states."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"images/ala2-deeplda.png\" width=\"500\" /></center>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Aldol reaction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this second example we consider the second example reported in the paper, which is an intermolecular aldol reaction between vinyl alcohol and formaldehyde.\n",
    "\n",
    "We download the data from the [DeepLDA repository](https://github.com/luigibonati/data-driven-CVs/) associated to the [PLUMED-NEST](https://www.plumed-nest.org/eggs/20/004/) submission, which contains unbiased simulations of the reactants and products.\n",
    "\n",
    "<center><img src=\"images/aldol-R.jpg\" width=\"500\" /><img src=\"images/aldol-P.jpg\" width=\"500\" /></center>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class 0 dataframe shape:  (5001, 43)\n",
      "Class 1 dataframe shape:  (5001, 43)\n",
      "\n",
      " - Loaded dataframe (10002, 43): ['time', 'cc2.0', 'cc2.1', 'cc2.2', 'oo2.0', 'co2.0', 'co2.1', 'co2.2', 'co2.3', 'co2.4', 'co2.5', 'ch2.0', 'ch2.1', 'ch2.2', 'ch2.3', 'ch2.4', 'ch2.5', 'ch2.6', 'ch2.7', 'ch2.8', 'ch2.9', 'ch2.10', 'ch2.11', 'ch2.12', 'ch2.13', 'ch2.14', 'ch2.15', 'ch2.16', 'ch2.17', 'oh2.0', 'oh2.1', 'oh2.2', 'oh2.3', 'oh2.4', 'oh2.5', 'oh2.6', 'oh2.7', 'oh2.8', 'oh2.9', 'oh2.10', 'oh2.11', 'walker', 'labels']\n",
      " - Descriptors (10002, 40): ['cc2.0', 'cc2.1', 'cc2.2', 'oo2.0', 'co2.0', 'co2.1', 'co2.2', 'co2.3', 'co2.4', 'co2.5', 'ch2.0', 'ch2.1', 'ch2.2', 'ch2.3', 'ch2.4', 'ch2.5', 'ch2.6', 'ch2.7', 'ch2.8', 'ch2.9', 'ch2.10', 'ch2.11', 'ch2.12', 'ch2.13', 'ch2.14', 'ch2.15', 'ch2.16', 'ch2.17', 'oh2.0', 'oh2.1', 'oh2.2', 'oh2.3', 'oh2.4', 'oh2.5', 'oh2.6', 'oh2.7', 'oh2.8', 'oh2.9', 'oh2.10', 'oh2.11']\n"
     ]
    }
   ],
   "source": [
    "# Set seed for reproducibility\n",
    "torch.manual_seed(1)\n",
    "\n",
    "from mlcolvar.io import create_dataset_from_files\n",
    "from mlcolvar.data import DictModule\n",
    "\n",
    "filenames = [ \"https://raw.githubusercontent.com/luigibonati/data-driven-CVs/master/aldol/1_unbiased/INPUTS.R\", \n",
    "              \"https://raw.githubusercontent.com/luigibonati/data-driven-CVs/master/aldol/1_unbiased/INPUTS.P\" ]\n",
    "\n",
    "n_states = len(filenames)\n",
    "\n",
    "# load dataset\n",
    "dataset, df = create_dataset_from_files(filenames,\n",
    "                                        filter_args={'regex':'cc|oo|co|ch|oh' }, # select contacts \n",
    "                                        create_labels=True,\n",
    "                                        return_dataframe=True, \n",
    "                                        )\n",
    "\n",
    "datamodule = DictModule(dataset,lengths=[0.8,0.2])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train DeepLDA CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DeepLDA(\n",
       "  (loss_fn): ReduceEigenvaluesLoss()\n",
       "  (norm_in): Normalization(in_features=40, out_features=40, mode=mean_std)\n",
       "  (nn): FeedForward(\n",
       "    (nn): Sequential(\n",
       "      (0): Linear(in_features=40, out_features=30, bias=True)\n",
       "      (1): Shifted_Softplus(beta=1, threshold=20)\n",
       "      (2): Linear(in_features=30, out_features=30, bias=True)\n",
       "      (3): Shifted_Softplus(beta=1, threshold=20)\n",
       "      (4): Linear(in_features=30, out_features=5, bias=True)\n",
       "    )\n",
       "  )\n",
       "  (lda): LDA(in_features=5, out_features=1)\n",
       ")"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from mlcolvar.cvs import DeepLDA\n",
    "\n",
    "n_input           = dataset['data'].shape[-1]\n",
    "nodes             = [n_input,30,30,5]\n",
    "nn_args           = {'activation': 'shifted_softplus'}\n",
    "options           = {'nn': nn_args}\n",
    "# MODEL\n",
    "model = DeepLDA(nodes, n_states=n_states, options=options)\n",
    "\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "GPU available: True (cuda), used: True\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/lightning/pytorch/loops/utilities.py:70: PossibleUserWarning: `max_epochs` was not set. Setting it to 1000 epochs. To train without an epoch limit, set `max_epochs=-1`.\n",
      "  rank_zero_warn(\n",
      "You are using a CUDA device ('NVIDIA GeForce RTX 3090') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision\n",
      "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0]\n",
      "\n",
      "  | Name    | Type                  | Params | In sizes | Out sizes\n",
      "-------------------------------------------------------------------------\n",
      "0 | loss_fn | ReduceEigenvaluesLoss | 0      | ?        | ?        \n",
      "1 | norm_in | Normalization         | 0      | [40]     | [40]     \n",
      "2 | nn      | FeedForward           | 2.3 K  | [40]     | [5]      \n",
      "3 | lda     | LDA                   | 0      | [5]      | [1]      \n",
      "-------------------------------------------------------------------------\n",
      "2.3 K     Trainable params\n",
      "0         Non-trainable params\n",
      "2.3 K     Total params\n",
      "0.009     Total estimated model params size (MB)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                                            "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/lbonati@iit.local/software/anaconda3/envs/pytorch/lib/python3.10/site-packages/lightning/pytorch/loops/fit_loop.py:280: PossibleUserWarning: The number of training batches (1) is smaller than the logging interval Trainer(log_every_n_steps=50). Set a lower value for log_every_n_steps if you want to see logs for the training epoch.\n",
      "  rank_zero_warn(\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 566: 100%|██████████| 1/1 [00:00<00:00, 19.76it/s, v_num=73] \n"
     ]
    }
   ],
   "source": [
    "from lightning.pytorch.callbacks.early_stopping import EarlyStopping\n",
    "from mlcolvar.utils.trainer import MetricsCallback\n",
    "\n",
    "# define callbacks\n",
    "metrics = MetricsCallback()\n",
    "early_stopping = EarlyStopping(monitor=\"valid_loss\", min_delta=0.1, patience=50)\n",
    "\n",
    "# define trainer\n",
    "trainer = lightning.Trainer(callbacks=[metrics, early_stopping],\n",
    "                     max_epochs=None, logger=None, enable_checkpointing=False)\n",
    "\n",
    "# fit\n",
    "trainer.fit( model, datamodule )"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Monitor learning curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mlcolvar.utils.plot import plot_metrics\n",
    "\n",
    "ax = plot_metrics(metrics.metrics, \n",
    "                  keys=[x for x in  metrics.metrics.keys() if 'eigval' in x and 'epoch' in x],#['train_loss_epoch','valid_loss'],\n",
    "                  yscale='linear')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Analysis of the CV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# evaluate cv on training set and append to dataframe\n",
    "X = dataset[:]['data']\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    s = model(torch.Tensor(X)).numpy()\n",
    "\n",
    "n_components = n_states-1\n",
    "\n",
    "for i in range(n_components):\n",
    "    df[f'CV{i}'] = s[:,i] "
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can evaluate the histogram of the CV on the training set, from which we see how the two states are well discriminated. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,axs = plt.subplots(1,n_components,figsize=(5*n_components,4))\n",
    "if n_components == 1:\n",
    "    axs = [axs]\n",
    "\n",
    "for i,ax in enumerate(axs):\n",
    "    ax.hist(s[:,i],bins=100)\n",
    "    ax.set_xlabel(f'CV {i}')\n",
    "    ax.set_ylabel('Histogram')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Features ranking"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can compute the feature importances by performing a sensitivity analysis, which amounts as measuring how sensitive is the CV $s$ to changes in the input features $x_i$, over a dataset $\\{\\mathbf{x}^{(j)}\\}_{j=1} ^N$:\n",
    "\n",
    "$$s_i = \\frac{1}{N} \\sum_j \\left|{\\frac{\\partial s}{\\partial x_i}(\\mathbf{x}^{(j)})}\\right| \\sigma_i$$\n",
    "\n",
    "where $\\sigma_i$ is the standard deviation, which is needed if the dataset is not standardized."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x1000 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mlcolvar.explain import sensitivity_analysis\n",
    "\n",
    "results = sensitivity_analysis(model, dataset, feature_names=dataset.feature_names, per_class=False, plot_mode='barh')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This allows us to identify three dominant features. Indeed, if we plot their distributions, we find that these features undergo significant changes when going from the reactant to the product states"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x300 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_n_features = 3\n",
    "\n",
    "names = results['feature_names'][-plot_n_features:]\n",
    "sensitivity = results['sensitivity']['Dataset'][-plot_n_features:]\n",
    "\n",
    "fig, axs = plt.subplots(1,plot_n_features,figsize=(3*plot_n_features,3))\n",
    "plt.suptitle('Distribution of the most relevant features')\n",
    "\n",
    "for ax,name,sensitivity in zip(axs.flatten()[::-1],names,sensitivity):\n",
    "    df.pivot(columns='labels')[name].plot.hist(bins=50,alpha=0.5,ax=ax)\n",
    "    ax.set_title(f'{name} ({np.round(sensitivity*100)}%)')\n",
    "    ax.legend(title='States',frameon=False)\n",
    "\n",
    "plt.tight_layout()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "pytorch13",
   "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": "3.10.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}