examples/02_quick_start.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "HyperTS is the DataCanvas Automatic Toolkit (DAT) tool chain, relying on [Hypernets](https://github.com/DataCanvasIO/Hypernets)-derived full Pipeline automatic toolkit for time series analysis. It follows the usage of ```make_expriment``` (similar to the API of [HyperGBM](https://github.com/DataCanvasIO/HyperGBM), an AutoML toolit for structured tabular data), and also conforms to the usage specification of the model API in ```scikit-learn```. We can create a ```make_expriment```, ```run``` and get a pipeline_model, which is a final optimized estimator, and then use its ```predict```, ```evaluate```, and ```plot``` to analyze unknown data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "HyperTS can be used to solve time series forecasting, classification and regression tasks, and they share a unified API. Next, we will divide into a quick demonstration of how to use it for time series forecasting and classification tasks."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Forecast Case"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.1 Prepare Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have doubts about the data format of HyperTS, please refer to the introduction in [01_datatypes_for_hyperts.ipynb](https://github.com/DataCanvasIO/HyperTS/blob/main/examples/01_datatypes_for_hyperts.ipynb)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts.datasets import load_network_traffic\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = load_network_traffic()\n",
    "train_data, test_data = train_test_split(df, test_size=168, shuffle=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Due to the temporal order of the data, in order to prevent information leakage, we split a part from the back of the overall data set, so ```shuffle=False```."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case dataset, we expose some basic information for reference, as follows:\n",
    "\n",
    "- time column: 'TimeStamp';\n",
    "- target columns: ['Var_1', 'Var_2', 'Var_3', 'Var_4', 'Var_5', 'Var_6'];\n",
    "- covariate columns: ['HourSin', 'WeekCos', 'CBWD'];\n",
    "- freq: 'H'."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.2 Create Experiment and Run"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We search for a time series model by creating ```make_experiment```, and call the ```run()``` method to perform the experiment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts import make_experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "experiment = make_experiment(train_data=train_data.copy(),\n",
    "                             task='forecast',\n",
    "                             mode='dl',\n",
    "                             timestamp='TimeStamp',\n",
    "                             covariables=['HourSin', 'WeekCos', 'CBWD'])\n",
    "model = experiment.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "In the forecasting task, we must pass the parameter ```timestamp``` column name to ```make_experiment```. If there are covariates, we also need to pass in the covariables column name.\n",
    "\n",
    "Therefore, in this case, we need to pass the following parameters to ```make_experiment```:\n",
    "\n",
    "1. Task type, ```task='forecast'```;\n",
    "2. Time column name, ```timestamp='TimeStamp'```;\n",
    "3. Covariate column name (ignore if not), ```covariables=['HourSin', 'WeekCos', 'CBWD']```;\n",
    "4. If you want to obtain powerful performance, you can also modify other default parameters. For details, please refer to the parameter introduction in ```make_experiment```."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.3 Print Pipeline Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<bound method Pipeline.get_params of Pipeline(steps=[('data_preprocessing',\n",
       "                 TSFDataPreprocessStep(covariate_cols=['HourSin', 'WeekCos',\n",
       "                                                       'CBWD'],\n",
       "                                       covariate_data_clean_args={'correct_object_dtype': False,\n",
       "                                                                  'drop_columns': None,\n",
       "                                                                  'drop_constant_columns': True,\n",
       "                                                                  'drop_duplicated_columns': False,\n",
       "                                                                  'drop_idness_columns': True,\n",
       "                                                                  'drop_label_nan_rows': True,\n",
       "                                                                  'int_convert_to': 'float',\n",
       "                                                                  'nan_chars': None,\n",
       "                                                                  'reduce_mem_usage': False,\n",
       "                                                                  'reserve_columns': None},\n",
       "                                       covariate_data_cleaner_args={'correct_object_dtype': False},\n",
       "                                       freq='H', name='data_preprocessing',\n",
       "                                       timestamp_col=['TimeStamp'])),\n",
       "                ('estimator',\n",
       "                 <hyperts.hyper_ts.HyperTSEstimator object at 0x000001885D81CAC8>)])>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_pipeline_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.4 Infer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split the test data into X and y, and call the ```predict()``` method to infer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>TimeStamp</th>\n",
       "      <th>Var_1</th>\n",
       "      <th>Var_2</th>\n",
       "      <th>Var_3</th>\n",
       "      <th>Var_4</th>\n",
       "      <th>Var_5</th>\n",
       "      <th>Var_6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2710</th>\n",
       "      <td>2021-06-23 19:00:00</td>\n",
       "      <td>1.580599</td>\n",
       "      <td>5.803027</td>\n",
       "      <td>6.339262</td>\n",
       "      <td>1.997845</td>\n",
       "      <td>68466.254953</td>\n",
       "      <td>3.615938e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2711</th>\n",
       "      <td>2021-06-23 20:00:00</td>\n",
       "      <td>2.564713</td>\n",
       "      <td>6.819843</td>\n",
       "      <td>19.643646</td>\n",
       "      <td>1.815659</td>\n",
       "      <td>25322.043751</td>\n",
       "      <td>2.768165e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2712</th>\n",
       "      <td>2021-06-23 21:00:00</td>\n",
       "      <td>2.436701</td>\n",
       "      <td>6.414492</td>\n",
       "      <td>11.427074</td>\n",
       "      <td>2.403198</td>\n",
       "      <td>268767.957348</td>\n",
       "      <td>8.900508e+05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2713</th>\n",
       "      <td>2021-06-23 22:00:00</td>\n",
       "      <td>2.444191</td>\n",
       "      <td>5.274575</td>\n",
       "      <td>9.378210</td>\n",
       "      <td>2.924055</td>\n",
       "      <td>53885.514664</td>\n",
       "      <td>1.774984e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2714</th>\n",
       "      <td>2021-06-23 23:00:00</td>\n",
       "      <td>2.276460</td>\n",
       "      <td>6.358713</td>\n",
       "      <td>7.335743</td>\n",
       "      <td>2.060555</td>\n",
       "      <td>118858.325416</td>\n",
       "      <td>9.228722e+05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               TimeStamp     Var_1     Var_2      Var_3     Var_4  \\\n",
       "2710 2021-06-23 19:00:00  1.580599  5.803027   6.339262  1.997845   \n",
       "2711 2021-06-23 20:00:00  2.564713  6.819843  19.643646  1.815659   \n",
       "2712 2021-06-23 21:00:00  2.436701  6.414492  11.427074  2.403198   \n",
       "2713 2021-06-23 22:00:00  2.444191  5.274575   9.378210  2.924055   \n",
       "2714 2021-06-23 23:00:00  2.276460  6.358713   7.335743  2.060555   \n",
       "\n",
       "              Var_5         Var_6  \n",
       "2710   68466.254953  3.615938e+05  \n",
       "2711   25322.043751  2.768165e+05  \n",
       "2712  268767.957348  8.900508e+05  \n",
       "2713   53885.514664  1.774984e+06  \n",
       "2714  118858.325416  9.228722e+05  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test, y_test = model.split_X_y(test_data.copy())\n",
    "forecast = model.predict(X_test)\n",
    "forecast.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.5 Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```evaluate``` method to perform the result evaluation, and we can observe the score under each evaluation indicator.\n",
    "\n",
    "Some default metrics scores will be returned here. If you want to observe the scores of the specified metrics, you can set the parameter metrics, such as ```metrics=['mae', 'mse', mape_func]```.\n",
    "\n",
    "Where, mape_func can be a custom evaluation function or an evaluation function from sklearn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Metirc</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>mae</td>\n",
       "      <td>143558.9729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>mse</td>\n",
       "      <td>179856088528.9392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rmse</td>\n",
       "      <td>194672.5783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>mape</td>\n",
       "      <td>0.5269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>smape</td>\n",
       "      <td>0.5107</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Metirc             Score\n",
       "0    mae       143558.9729\n",
       "1    mse 179856088528.9392\n",
       "2   rmse       194672.5783\n",
       "3   mape            0.5269\n",
       "4  smape            0.5107"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = model.evaluate(y_true=y_test, y_pred=forecast)\n",
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.6 Visualization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```plot()``` method to visualize, take a look at the forecast curve, and compare it with the actual curve."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "model.plot(forecast=forecast, actual=test_data, var_id='Var_3', interactive=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "- Only the forecat curve of one variable will be displayed here, which defaults to the first target variable;\n",
    "- For multivariate forecasting task, if we want to observe other target variable curve, you can modify the parameter ```var_id```. For example: var_id=2 (i.e., index) or var_id='Var_2' (i.e., name);\n",
    "- ```plot``` supports interactive visualization by changing ```interactive=False```(default， need to install plotly);\n",
    "- To plot longer-term historical information, set the parameter ```history=sub_train_data```."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.7 Save model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```save()``` method to save the trained model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(model_file=\"./xxx/xxx/models\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.8 Load model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```load_model()``` method to load the saved model. Note the path."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts.utils.models import load_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pipeline_model = load_model(model_file=\"./xxx/xxx/models/stats_models\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Classification Case"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.1 Prepare Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If you have doubts about the data format of HyperTS, please refer to the introduction in [01_datatypes_for_hyperts.ipynb](https://github.com/DataCanvasIO/HyperTS/blob/main/examples/01_datatypes_for_hyperts.ipynb)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts.datasets import load_basic_motions\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = load_basic_motions()\n",
    "train_df, test_df = train_test_split(df, test_size=0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this case dataset, we expose some basic information for reference, as follows:\n",
    "\n",
    "- feature columns: ['Var_1', 'Var_2', 'Var_3', 'Var_4', 'Var_5', 'Var_6'];\n",
    "- target column: 'target'."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.2 Create Experiment and Run"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We search for a time series model by creating ```make_experiment```, and call the ```run()``` method to perform the experiment."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts import make_experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "experiment = make_experiment(train_data=train_df.copy(), task='classification', mode='stats', target='target')\n",
    "\n",
    "model = experiment.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.3 Print Pipeline Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<bound method Pipeline.get_params of Pipeline(steps=[('data_preprocessing',\n",
       "                 TSCDataPreprocessStep(data_cleaner_args={'drop_label_nan_rows': True},\n",
       "                                       name='data_preprocessing')),\n",
       "                ('estimator',\n",
       "                 <hyperts.hyper_ts.HyperTSEstimator object at 0x000001CDC2318C18>)])>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_pipeline_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.4 Infer"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Split the test data into X and y, and call the ```predict()``` method to infer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test, y_test = model.split_X_y(test_df.copy())\n",
    "y_pred = model.predict(X_test)\n",
    "y_proba = model.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['badminton', 'walking', 'standing', 'badminton', 'walking',\n",
       "       'walking', 'walking', 'badminton', 'standing', 'running',\n",
       "       'badminton', 'walking', 'walking', 'running', 'standing',\n",
       "       'standing'], dtype=object)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.5 Evaluation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```evaluate``` method to perform the result evaluation, and we can observe the score under each evaluation indicator.\n",
    "\n",
    "Some default metrics scores will be returned here. If you want to observe the scores of the specified metrics, you can set the parameter metrics, such as ```metrics=['accuracy', 'auc', f1_func]```.\n",
    "\n",
    "Where, mape_func can be a custom evaluation function or an evaluation function from sklearn."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>Metirc</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>accuracy</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>f1</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>precision</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>recall</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Metirc  Score\n",
       "0   accuracy 1.0000\n",
       "1         f1 1.0000\n",
       "2  precision 1.0000\n",
       "3     recall 1.0000"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = model.evaluate(y_true=y_test, y_pred=y_pred, y_proba=y_proba)\n",
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1.6 Save model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Call the ```save()``` method to save the trained model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hyperts.utils.models import save_model\n",
    "\n",
    "save_model(model=model, model_file=\"./xxx/xxx/models\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}