cif_VVN.ipynb

{
  "cells": [
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "_v7BmDsQgutC"
      },
      "source": [
        "## Predicting $\\Gamma$-phonon with VVN given CIF files\n",
        "### Importing Required Libraries and Configuring the Environment"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {},
      "outputs": [],
      "source": [
        "import os\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib as mpl\n",
        "import torch\n",
        "from torch_geometric.loader import DataLoader\n",
        "import pandas as pd\n",
        "from ase.visualize.plot import plot_atoms\n",
        "from ase.io import read\n",
        "from utils.utils_data import generate_data_dict\n",
        "from utils.utils_model import GraphNetwork_VVN\n",
        "from utils.utils_train import load_model\n",
        "from utils.utils_plot import generate_dataframe, plot_gphonon\n",
        "from utils.utils_qpts import get_qpts\n",
        "from utils.helpers import sub\n",
        "from config_file import pretrained_model_dict, palette\n",
        "torch.set_default_dtype(torch.float64)\n",
        "device = 'cuda' if torch.cuda.is_available() else 'cpu'\n",
        "color_set = ['g', 'y', 'r']\n",
        "colors = dict(zip(color_set, palette[:len(color_set)]))\n",
        "cmap = mpl.colors.LinearSegmentedColormap.from_list('cmap', [palette[k] for k in [0,2,1]])"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "###  Setting Model Parameters"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [],
      "source": [
        "option = 'vvn' # use kMVN model\n",
        "descriptor = 'mass' # use mass as descriptor\n",
        "use_pretrained = True\n",
        "pretrained_name = pretrained_model_dict[option] # pretrained model uses 'mass' as descriptor\n",
        "r_max = 4.0 # cutoff radius (we used 4.0 Angstrom in the paper)\n",
        "factor = 1000   # factor to scale the predicted band gap (do not need to change)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Loading CIF Files and Preparing Data\n",
        "We search cif files in cif_folder, and load them with high-symmetry momentum path suggested by SeekPath. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/home/rokabe/anaconda3/envs/phonon/lib/python3.9/site-packages/seekpath/hpkot/__init__.py:286: EdgeCaseWarning: hR lattice, but sqrt(3)a almost equal to sqrt(2)c\n",
            "  warnings.warn(\n",
            "/home/rokabe/anaconda3/envs/phonon/lib/python3.9/site-packages/seekpath/hpkot/__init__.py:352: EdgeCaseWarning: aP lattice, but the k_alpha3 angle is almost equal to 90 degrees\n",
            "  warnings.warn(\n",
            "/home/rokabe/anaconda3/envs/phonon/lib/python3.9/site-packages/seekpath/hpkot/__init__.py:357: EdgeCaseWarning: aP lattice, but the k_beta3 angle is almost equal to 90 degrees\n",
            "  warnings.warn(\n",
            "/home/rokabe/anaconda3/envs/phonon/lib/python3.9/site-packages/seekpath/hpkot/__init__.py:362: EdgeCaseWarning: aP lattice, but the k_gamma3 angle is almost equal to 90 degrees\n",
            "  warnings.warn(\n",
            "  0%|          | 0/5 [00:00<?, ?it/s]/data1/rokabe/phonon/phonon_prediction/utils/utils_data.py:267: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
            "  'edge_shift': torch.tensor(edge_shift, dtype = torch.float64), 'edge_vec': torch.tensor(edge_vec, dtype = torch.float64),\n",
            "100%|██████████| 5/5 [00:00<00:00, 98.03it/s]\n"
          ]
        }
      ],
      "source": [
        "cif_folder = './cif_folder'\n",
        "cif_files = sorted([f for f in os.listdir(cif_folder) if f.endswith('.cif')])\n",
        "data = pd.DataFrame({})\n",
        "for cif_file in cif_files:\n",
        "    d = dict()\n",
        "    atoms = read(os.path.join(cif_folder, cif_file))\n",
        "    material_id = cif_file.replace('.cif', '')\n",
        "    numb = len(atoms)\n",
        "\n",
        "    qpts_dict = get_qpts(atoms)\n",
        "    qpts = np.array(qpts_dict['qpts'])\n",
        "    qticks = qpts_dict['qticks']\n",
        "    d['id'] = material_id\n",
        "    d['structure'] = [atoms]\n",
        "    d['qpts'] = [qpts]\n",
        "    d['qticks'] = [qticks]\n",
        "    d['real_band'] = [np.zeros((len(qpts), 3*numb))]    # Since we do not have band structure data, we set it to zero matrix of shape (len(qpts), 3*numb)\n",
        "    dfn = pd.DataFrame(data = d)\n",
        "    data = pd.concat([data, dfn], ignore_index = True)\n",
        "\n",
        "data_dict = generate_data_dict(data, r_max, descriptor=descriptor, option=option, factor=factor)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>structure</th>\n",
              "      <th>qpts</th>\n",
              "      <th>qticks</th>\n",
              "      <th>real_band</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>mp-149</td>\n",
              "      <td>(Atom('Si', [0.0, 0.0, 0.0], index=0), Atom('S...</td>\n",
              "      <td>[[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...</td>\n",
              "      <td>[\\Gamma, , , , , , , , , , , , , , , , , T, H_...</td>\n",
              "      <td>[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>mp-2059</td>\n",
              "      <td>(Atom('Nb', [0.0, 0.0, 8.8542772787448], index...</td>\n",
              "      <td>[[0.0, 0.0, 0.0], [0.05555555555555555, 0.0, 0...</td>\n",
              "      <td>[\\Gamma, , , , , , , , , X, Y, , , , , , , , ,...</td>\n",
              "      <td>[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>mp-2534</td>\n",
              "      <td>(Atom('Ga', [0.0, 0.0, 0.0], index=0), Atom('G...</td>\n",
              "      <td>[[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...</td>\n",
              "      <td>[\\Gamma, , , , , , , , , , , , , , , , , T, H_...</td>\n",
              "      <td>[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>mp-30</td>\n",
              "      <td>(Atom('Cu', [0.0, 0.0, 0.0], index=0), Atom('C...</td>\n",
              "      <td>[[0.0, 0.0, 0.0], [0.0, 0.05555555555555555, 0...</td>\n",
              "      <td>[\\Gamma, , , , , , , , , X, , , , , , , , , M,...</td>\n",
              "      <td>[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>mp-7577</td>\n",
              "      <td>(Atom('Co', [0.639511351248549, 0.639511351248...</td>\n",
              "      <td>[[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...</td>\n",
              "      <td>[\\Gamma, , , , , , , , , , , , , , , , , T, H_...</td>\n",
              "      <td>[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        id                                          structure  \\\n",
              "0   mp-149  (Atom('Si', [0.0, 0.0, 0.0], index=0), Atom('S...   \n",
              "1  mp-2059  (Atom('Nb', [0.0, 0.0, 8.8542772787448], index...   \n",
              "2  mp-2534  (Atom('Ga', [0.0, 0.0, 0.0], index=0), Atom('G...   \n",
              "3    mp-30  (Atom('Cu', [0.0, 0.0, 0.0], index=0), Atom('C...   \n",
              "4  mp-7577  (Atom('Co', [0.639511351248549, 0.639511351248...   \n",
              "\n",
              "                                                qpts  \\\n",
              "0  [[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...   \n",
              "1  [[0.0, 0.0, 0.0], [0.05555555555555555, 0.0, 0...   \n",
              "2  [[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...   \n",
              "3  [[0.0, 0.0, 0.0], [0.0, 0.05555555555555555, 0...   \n",
              "4  [[0.0, 0.0, 0.0], [0.029411764705882353, 0.029...   \n",
              "\n",
              "                                              qticks  \\\n",
              "0  [\\Gamma, , , , , , , , , , , , , , , , , T, H_...   \n",
              "1  [\\Gamma, , , , , , , , , X, Y, , , , , , , , ,...   \n",
              "2  [\\Gamma, , , , , , , , , , , , , , , , , T, H_...   \n",
              "3  [\\Gamma, , , , , , , , , X, , , , , , , , , M,...   \n",
              "4  [\\Gamma, , , , , , , , , , , , , , , , , T, H_...   \n",
              "\n",
              "                                           real_band  \n",
              "0  [[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  \n",
              "1  [[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  \n",
              "2  [[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  \n",
              "3  [[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  \n",
              "4  [[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0,...  "
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Review the data\n",
        "data"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Transform the data into the input data format. \n",
        "Get the test data with the proper format. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [],
      "source": [
        "te_num = len(data_dict)\n",
        "idx_te = range(te_num)\n",
        "data_set = torch.utils.data.Subset(list(data_dict.values()), idx_te)\n",
        "te_set = torch.utils.data.Subset(data_set, idx_te)\n",
        "te_loader = DataLoader(te_set, batch_size = 1)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Load pre-trained model\n",
        "Load the pre-trained model from the archive. Our code automatically load the hyperparameters and the learned weights from the .torch file. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Loading model from: ./models/vvn_241026-180553.torch\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/home/rokabe/anaconda3/envs/phonon/lib/python3.9/site-packages/torch/jit/_check.py:172: UserWarning: The TorchScript type system doesn't support instance-level annotations on empty non-base types in `__init__`. Instead, either 1) use a type annotation in the class body, or 2) wrap the type in `torch.jit.Attribute`.\n",
            "  warnings.warn(\"The TorchScript type system doesn't support \"\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "number of parameters:  250328\n"
          ]
        }
      ],
      "source": [
        "\n",
        "model_name = pretrained_name # model name (load from pretrained model)\n",
        "save_name = f'./models/{model_name}'   \n",
        "model_file = save_name + '.torch'\n",
        "model, conf_dict, history, s0 = load_model(GraphNetwork_VVN, model_file, device)\n",
        "num_params = sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
        "print('number of parameters: ', num_params)"
      ]
    },
    {
      "attachments": {},
      "cell_type": "markdown",
      "metadata": {
        "id": "5kpYlnpwEFW3"
      },
      "source": [
        "### Predict the phonon band structure\n",
        "We use the pre-trained model to predict phonon of the given cif files. "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "  0%|          | 0/5 [00:00<?, ?it/s]/data1/rokabe/phonon/phonon_prediction/utils/utils_plot.py:168: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.\n",
            "  df = pd.concat([df, df0], ignore_index=True)\n",
            "100%|██████████| 5/5 [00:02<00:00,  1.80it/s]\n"
          ]
        }
      ],
      "source": [
        "# Generate Data Loader and then Data Frame\n",
        "df_te = generate_dataframe(model, te_loader, None, device, option, factor)  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, '[mp-149] Si₈: $\\\\Gamma$-phonon')"
            ]
          },
          "execution_count": 8,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 600x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 200x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Set the index for which structure to plot from the dataframe\n",
        "idx_out = 0  #TODO: change this index to plot a different structure (0~len(data)-1)\n",
        "\n",
        "# [0] Load the structure and other data for the selected material\n",
        "row_data = data.iloc[idx_out]  # Get the row for the selected structure in the data dataframe\n",
        "material_id = row_data['id']  # Extract the material ID\n",
        "struct = row_data['structure']  # Get the structure of the material from the dataframe\n",
        "row_df = df_te[df_te['id'] == material_id]  # Get the row for the selected structure in the test set\n",
        "pred = row_df['pred'].item()  # Extract the predicted band structure\n",
        "real = None  # Set real data to None if not available\n",
        "symbols = np.unique(list(struct.symbols))  # Get unique atomic symbols from the structure\n",
        "formula = struct.get_chemical_formula().translate(sub)  # Get the chemical formula of the structure\n",
        "numb = len(struct)  # Get the number of atoms in the structure\n",
        "z = dict(zip(symbols, range(numb)))  # Create a mapping from atomic symbols to indices\n",
        "\n",
        "# [1] Create a plot of the atomic structure using matplotlib\n",
        "fig, ax = plt.subplots(figsize=(6,6))\n",
        "norm = plt.Normalize(vmin=0, vmax=numb-1)  # Normalize the color map for atoms\n",
        "color = [mpl.colors.to_hex(k) for k in cmap(norm([z[j] for j in list(struct.symbols)]))]  # Assign colors based on atomic symbols\n",
        "plot_atoms(struct, ax, radii=0.25, colors=color, rotation=('30x,30y,0z'))  # Plot the atoms with radii and rotation\n",
        "ax.set_xlabel(r'$x_1\\ (\\AA)$')  # Set x-axis label with LaTeX formatting for Angstroms\n",
        "ax.set_ylabel(r'$x_2\\ (\\AA)$')  # Set y-axis label with LaTeX formatting\n",
        "ax.set_title(f\"[{material_id}] {formula}: crystal structure\")  \n",
        "\n",
        "# [2] Plot the predicted phonon band structure\n",
        "# Plot the predicted band structure with custom function\n",
        "fig, ax = plt.subplots(figsize=(2,6))\n",
        "plot_gphonon(ax, real, pred, palette[-1], lwidth=0.8, plot_real=False, ylabel=True)  # Plot prediction without real data\n",
        "ax.set_title(f\"[{material_id}] {formula}: $\\Gamma$-phonon\")  \n"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "authorship_tag": "ABX9TyP9WdY8pU0J3fdzaTpnaMJ8",
      "collapsed_sections": [],
      "include_colab_link": true,
      "name": "phononDoS_colab.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3.9.13 ('pdos': conda)",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.13"
    },
    "vscode": {
      "interpreter": {
        "hash": "e1f4bc6f35d515ca6a525bb4b11c3080afe8136d3216b199fcb3aa4a51b71c9c"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}