{
 "cells": [
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "# Saving and loading models\n",
    "\n",
    "This notebook introduces the concept of saving and loading `.gwref` files.\n",
    "\n",
    "This notebook can be downloaded from the source code [here](https://github.com/andersretznerSGU/gwrefpy/blob/main/docs/user_guide/5_io.ipynb).\n",
    "\n",
    "# Saving a model 💾\n",
    "\n",
    "You can save a `gwrefpy` model to a `.gwref` file using the `save_project` function of the `Model` class. This function saves the model, including all wells and fits, to a specified file."
   ],
   "id": "f8ebfe5589158bd0"
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "We will first create a simple model with one observed well and one reference well.",
   "id": "a7d320daa081a770"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-19T06:19:23.202275Z",
     "start_time": "2025-09-19T06:19:21.423475Z"
    }
   },
   "cell_type": "code",
   "source": [
    "import gwrefpy as gr\n",
    "import numpy as np\n",
    "import pandas as pd"
   ],
   "id": "ae7b6686cad2a72f",
   "outputs": [],
   "execution_count": 3
  },
  {
   "metadata": {
    "tags": [
     "hide-input"
    ],
    "ExecuteTime": {
     "end_time": "2025-09-19T06:19:23.878Z",
     "start_time": "2025-09-19T06:19:23.865523Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Create some example data\n",
    "n_days = 100\n",
    "dates = pd.date_range(\"2020-01-01\", periods=n_days, freq=\"D\")\n",
    "\n",
    "# Observed and reference values with some noise\n",
    "values_obs1 = (\n",
    "    25.75\n",
    "    + 0.7 * np.sin(np.linspace(0, 4 * np.pi, n_days))\n",
    "    + np.random.normal(0, 0.1, n_days)\n",
    ")\n",
    "values_obs1 = pd.Series(values_obs1, index=dates)\n",
    "values_ref1 = (\n",
    "    18.75\n",
    "    + 0.3 * np.sin(np.linspace(0, 4 * np.pi, n_days))\n",
    "    + np.random.normal(0, 0.05, n_days)\n",
    ")\n",
    "values_ref1 = pd.Series(values_ref1, index=dates)"
   ],
   "id": "e9337241bbe35e1a",
   "outputs": [],
   "execution_count": 4
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-19T06:19:25.524811Z",
     "start_time": "2025-09-19T06:19:25.522288Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Creat the observed and reference wells\n",
    "obs1 = gr.Well(name=\"Obs. well\", is_reference=False)\n",
    "obs1.add_timeseries(values_obs1)\n",
    "ref1 = gr.Well(name=\"Ref. well\", is_reference=True)\n",
    "ref1.add_timeseries(values_ref1)\n",
    "\n",
    "# Create the model and add the wells\n",
    "model1 = gr.Model(name=\"IO Model\")\n",
    "model1.add_well(obs1)\n",
    "model1.add_well(ref1)"
   ],
   "id": "dd1b89e7b780e90d",
   "outputs": [],
   "execution_count": 5
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "We can now save the model to a `.gwref` file using the `save_project` function.\n",
    "```{note}\n",
    "The funciton will create the file in the current working directory if no path is specified.\n",
    "```"
   ],
   "id": "3f437ef415865dca"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-19T06:21:35.698499Z",
     "start_time": "2025-09-19T06:21:35.680754Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Save the model to a .gwref file\n",
    "model1.save_project(\"my_model.gwref\")"
   ],
   "id": "cff5a365dc580981",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model 'Logging Example Model' saved to 'example_model.gwref'.\n"
     ]
    }
   ],
   "execution_count": 9
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "You can now find the `my_model.gwref` files in your working directory. You can also give a path with the filename and the fild will be saved to that path, e.g. `\"path//to//folder//my_model.gwref\"`. \n",
    "```{warning}\n",
    "If the file already exists, a warning will be logged and the file will not be overwritten. You can set the ``overwrite`` variable to `True` to overwrite the existing file.\n",
    "\n",
    "The funciton will not create any folders in the specified path if they do not exist.\n",
    "```"
   ],
   "id": "d2a3d40fe6011460"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-19T06:21:50.470844Z",
     "start_time": "2025-09-19T06:21:50.464055Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Save the model to a .gwref file, allowing overwriting\n",
    "model1.save_project(\"my_model.gwref\") # This will not overwrite the existing file and a warning will be logged\n",
    "model1.save_project(\"my_model.gwref\", overwrite=True) # This will overwrite the existing file"
   ],
   "id": "8927630e5d539af5",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The file 'example_model.gwref' already exists. To overwrite the existing file set the argument 'overwrite' to True.\n",
      "Model 'Logging Example Model' saved to 'example_model.gwref'.\n"
     ]
    }
   ],
   "execution_count": 10
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": [
    "## Loading Models 📦\n",
    "You can load a model from a `.gwref` file using the `open_project` function from the `Model` class. This function reads the model from the specified file and saves it to the `Model` object. The model will include all wells and fits that were saved in the file.\n",
    "\n",
    "There are two ways to load a model:\n",
    "1. Create a new `Model` object and use the `open_project` method.\n",
    "2. Specify the `.gwref` file directly when creating a new `Model` object."
   ],
   "id": "ab0429e99fb8f13e"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-18T11:56:24.801814Z",
     "start_time": "2025-09-18T11:56:24.769652Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Method 1: Create a new Model object and load the project\n",
    "loaded_model1 = gr.Model(name=\"Name will be overwritten\")\n",
    "loaded_model1.open_project(\"my_model.gwref\")"
   ],
   "id": "30f2208061681fe",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model 'Logging Example Model' loaded from 'example_model.gwref'.\n"
     ]
    }
   ],
   "execution_count": 3
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-09-18T11:56:24.887100Z",
     "start_time": "2025-09-18T11:56:24.877035Z"
    }
   },
   "cell_type": "code",
   "source": [
    "# Method 2: Directly load the project into a new Model object\n",
    "loaded_model2 = gr.Model(\"my_model.gwref\")"
   ],
   "id": "5b2b524c40b0d03",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model 'Logging Example Model' loaded from 'example_model.gwref'.\n"
     ]
    }
   ],
   "execution_count": 4
  },
  {
   "metadata": {},
   "cell_type": "markdown",
   "source": "This concludes this notebook on saving and loading models in `gwrefpy`.",
   "id": "e30b422c68de2ecf"
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}