09 Optimizing parameters in a PCSE model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img style=\"float: right;\" src=\"data:image/png;base64,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\">\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Optimizing parameters in a PCSE crop model\n",
    "\n",
    "In this Jupyter notebook we will demonstrate how to optimize parameters in a PCSE model. In this example a set of synthetic observations of LAI will be used to optimize the TWDI (total initial dry weight) and SPAN (life span of leaves) parameters of the PCSE/WOFOST model using an optimization algorithm.\n",
    "\n",
    "We use the SUBPLEX optimization algorithm implemented in the [NLOPT](https://nlopt.readthedocs.io/en/latest/) optimization library. NLOPT can be a bit hard to install. Your best bet is to first try to install NLOPT through the conda package manager which is part of the [anaconda](https://www.anaconda.com) python distribution. If that fails (I have seen this happening particularly on windows), you can download the library from the site of Christoph Gohlke [here](https://www.lfd.uci.edu/~gohlke/pythonlibs/) who provides pre-compiled windows binaries for many python libraries. In the latter case you can install the package with `pip install <package.whl>`.\n",
    "\n",
    "Allard de Wit, January 2018 (allard.dewit@wur.nl)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## First some relevant imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "This notebook was built with:\n",
      "python version: 3.10.13 | packaged by conda-forge | (main, Oct 26 2023, 18:01:37) [MSC v.1935 64 bit (AMD64)]\n",
      "PCSE version: 6.0.6\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import sys\n",
    "import matplotlib\n",
    "matplotlib.style.use(\"ggplot\")\n",
    "import matplotlib.pyplot as plt\n",
    "import yaml\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import nlopt\n",
    "from itertools import product\n",
    "from progressbar import printProgressBar\n",
    "\n",
    "import pcse\n",
    "from pcse.models import Wofost72_PP\n",
    "from pcse.base import ParameterProvider\n",
    "from pcse.input import NASAPowerWeatherDataProvider, YAMLAgroManagementReader, YAMLCropDataProvider, WOFOST72SiteDataProvider\n",
    "from pcse.util import DummySoilDataProvider\n",
    "print(\"This notebook was built with:\")\n",
    "print(f\"python version: {sys.version}\")\n",
    "print(f\"PCSE version: {pcse.__version__}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Collect necessary input data for WOFOST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Weather data for Netherlands\n",
    "wdp = NASAPowerWeatherDataProvider(latitude=52, longitude=5)\n",
    "# Standard crop parameter library\n",
    "cropd = YAMLCropDataProvider()\n",
    "# We don't need soil for potential production, so we use dummy values\n",
    "soild = DummySoilDataProvider()\n",
    "# Some site parameters\n",
    "sited = WOFOST72SiteDataProvider(WAV=5)\n",
    "# Package everyting into a single parameter object\n",
    "params = ParameterProvider(cropdata=cropd, sitedata=sited, soildata=soild)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Here we define the agromanagement for sugar beet\n",
    "agro_yaml = \"\"\"\n",
    "- 2006-01-01:\n",
    "    CropCalendar:\n",
    "        crop_name: sugarbeet\n",
    "        variety_name: Sugarbeet_603\n",
    "        crop_start_date: 2006-03-31\n",
    "        crop_start_type: emergence\n",
    "        crop_end_date: 2006-10-20\n",
    "        crop_end_type: harvest\n",
    "        max_duration: 300\n",
    "    TimedEvents: null\n",
    "    StateEvents: null\n",
    "\"\"\"\n",
    "agro = yaml.safe_load(agro_yaml)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Run the simulation with the default parameters\n",
    "We use the standard WOFOST output as pseudo observations, so first we run the model with the default crop parameters and collect the output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>day</th>\n",
       "      <th>DVS</th>\n",
       "      <th>LAI</th>\n",
       "      <th>TAGP</th>\n",
       "      <th>TWSO</th>\n",
       "      <th>TWLV</th>\n",
       "      <th>TWST</th>\n",
       "      <th>TWRT</th>\n",
       "      <th>TRA</th>\n",
       "      <th>RD</th>\n",
       "      <th>SM</th>\n",
       "      <th>WWLOW</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>day</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2006-10-16</th>\n",
       "      <td>2006-10-16</td>\n",
       "      <td>2.182200</td>\n",
       "      <td>1.318570</td>\n",
       "      <td>25045.166727</td>\n",
       "      <td>14916.025792</td>\n",
       "      <td>5112.997882</td>\n",
       "      <td>5016.143053</td>\n",
       "      <td>3982.758431</td>\n",
       "      <td>0.082313</td>\n",
       "      <td>120.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2006-10-17</th>\n",
       "      <td>2006-10-17</td>\n",
       "      <td>2.187357</td>\n",
       "      <td>1.327063</td>\n",
       "      <td>25130.092963</td>\n",
       "      <td>14992.459404</td>\n",
       "      <td>5117.244194</td>\n",
       "      <td>5020.389365</td>\n",
       "      <td>3984.491619</td>\n",
       "      <td>0.078241</td>\n",
       "      <td>120.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2006-10-18</th>\n",
       "      <td>2006-10-18</td>\n",
       "      <td>2.193236</td>\n",
       "      <td>1.335364</td>\n",
       "      <td>25213.103257</td>\n",
       "      <td>15067.168669</td>\n",
       "      <td>5121.394709</td>\n",
       "      <td>5024.539880</td>\n",
       "      <td>3986.185707</td>\n",
       "      <td>0.051568</td>\n",
       "      <td>120.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2006-10-19</th>\n",
       "      <td>2006-10-19</td>\n",
       "      <td>2.200414</td>\n",
       "      <td>1.312543</td>\n",
       "      <td>25242.014171</td>\n",
       "      <td>15093.188491</td>\n",
       "      <td>5122.840255</td>\n",
       "      <td>5025.985425</td>\n",
       "      <td>3986.775725</td>\n",
       "      <td>0.046447</td>\n",
       "      <td>120.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2006-10-20</th>\n",
       "      <td>2006-10-20</td>\n",
       "      <td>2.207914</td>\n",
       "      <td>1.315999</td>\n",
       "      <td>25276.576594</td>\n",
       "      <td>15124.294672</td>\n",
       "      <td>5124.568376</td>\n",
       "      <td>5027.713546</td>\n",
       "      <td>3987.481081</td>\n",
       "      <td>0.067994</td>\n",
       "      <td>120.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   day       DVS       LAI          TAGP          TWSO  \\\n",
       "day                                                                      \n",
       "2006-10-16  2006-10-16  2.182200  1.318570  25045.166727  14916.025792   \n",
       "2006-10-17  2006-10-17  2.187357  1.327063  25130.092963  14992.459404   \n",
       "2006-10-18  2006-10-18  2.193236  1.335364  25213.103257  15067.168669   \n",
       "2006-10-19  2006-10-19  2.200414  1.312543  25242.014171  15093.188491   \n",
       "2006-10-20  2006-10-20  2.207914  1.315999  25276.576594  15124.294672   \n",
       "\n",
       "                   TWLV         TWST         TWRT       TRA     RD   SM WWLOW  \n",
       "day                                                                            \n",
       "2006-10-16  5112.997882  5016.143053  3982.758431  0.082313  120.0  0.3  None  \n",
       "2006-10-17  5117.244194  5020.389365  3984.491619  0.078241  120.0  0.3  None  \n",
       "2006-10-18  5121.394709  5024.539880  3986.185707  0.051568  120.0  0.3  None  \n",
       "2006-10-19  5122.840255  5025.985425  3986.775725  0.046447  120.0  0.3  None  \n",
       "2006-10-20  5124.568376  5027.713546  3987.481081  0.067994  120.0  0.3  None  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wofost = Wofost72_PP(params, wdp, agro)\n",
    "wofost.run_till_terminate()\n",
    "df = pd.DataFrame(wofost.get_output())\n",
    "df.index = pd.to_datetime(df.day)\n",
    "df.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sample the WOFOST output to generate some \"observations\"\n",
    "Let's assume that we have weekly field observations of LAI, therefore we resample the dataframe with WOFOST output."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# get daily observations for those\n",
    "ix = (df.index.dayofweek == 0) & (df.LAI.notnull())\n",
    "df_pseudo_obs = df.loc[ix]\n",
    "fig, axes = plt.subplots(figsize=(12,8))\n",
    "axes.plot_date(df_pseudo_obs.index, df_pseudo_obs.LAI)\n",
    "r = axes.set_title(\"Pseudo LAI observations\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define a model runner\n",
    "A model runner just reruns the WOFOST model with different values for a set of parameters, in this case \"TDWI\" and \"SPAN\". It then collects the model output, converts it to a dataframe and returns it. \n",
    "\n",
    "Changing the model parameters for each WOFOST simulation is done with the `set_override()` method on the parameter object."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ModelRerunner(object):\n",
    "    \"\"\"Reruns a given model with different values of parameters TWDI and SPAN.\n",
    "    \n",
    "    Returns a pandas DataFrame with simulation results of the model with given\n",
    "    parameter values.\n",
    "    \"\"\"\n",
    "    parameters = [\"TDWI\", \"SPAN\"]\n",
    "    \n",
    "    def __init__(self, params, wdp, agro):\n",
    "        self.params = params\n",
    "        self.wdp = wdp\n",
    "        self.agro = agro\n",
    "        \n",
    "    def __call__(self, par_values):\n",
    "        # Check if correct number of parameter values were provided\n",
    "        if len(par_values) != len(self.parameters):\n",
    "            msg = \"Optimizing %i parameters, but only % values were provided!\" % (len(self.parameters, len(par_values)))\n",
    "            raise RuntimeError(msg)\n",
    "        # Clear any existing overrides\n",
    "        self.params.clear_override()\n",
    "        # Set overrides for the new parameter values\n",
    "        for parname, value in zip(self.parameters, par_values):\n",
    "            self.params.set_override(parname, value)\n",
    "        # Run the model with given parameter values\n",
    "        wofost = Wofost72_PP(self.params, self.wdp, self.agro)\n",
    "        wofost.run_till_terminate()\n",
    "        df = pd.DataFrame(wofost.get_output())\n",
    "        df.index = pd.to_datetime(df.day)\n",
    "        return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define an object function calculator\n",
    "The objective function calculator runs the WOFOST model with the given set of input parameters, collects the simulation results and computes the differences with the observations. Different objective functions can be used, in this case we use the root of the mean squared error (RMSE)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ObjectiveFunctionCalculator(object):\n",
    "    \"\"\"Computes the objective function.\n",
    "    \n",
    "    This class runs the simulation model with given parameter values and returns the objective\n",
    "    function as the sum of squared difference between observed and simulated LAI.\n",
    ".   \"\"\"\n",
    "    \n",
    "    def __init__(self, params, wdp, agro, observations):\n",
    "        self.modelrerunner = ModelRerunner(params, wdp, agro)\n",
    "        self.df_observations = observations\n",
    "        self.n_calls = 0\n",
    "       \n",
    "    def __call__(self, par_values, grad=None):\n",
    "        \"\"\"Runs the model and computes the objective function for given par_values.\n",
    "        \n",
    "        The input parameter 'grad' must be defined in the function call, but is only\n",
    "        required for optimization methods where analytical gradients can be computed.\n",
    "        \"\"\"\n",
    "        self.n_calls += 1\n",
    "        print(\".\", end=\"\")\n",
    "        # Run the model and collect output\n",
    "        self.df_simulations = self.modelrerunner(par_values)\n",
    "        # compute the differences by subtracting the DataFrames\n",
    "        # Note that the dataframes automatically join on the index (dates) and column names\n",
    "        df_differences = self.df_simulations - self.df_observations\n",
    "        # Compute the RMSE on the LAI column\n",
    "        obj_func = np.sqrt(np.mean(df_differences.LAI**2))\n",
    "        return obj_func"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test the objective function calculator\n",
    "The objective function calculator should return a value of zero when called with the default \n",
    "values of TWDI and SPAN."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".Objective function value with default parameters ([0.51, 37.0]): 0.0\n"
     ]
    }
   ],
   "source": [
    "objfunc_calculator = ObjectiveFunctionCalculator(params, wdp, agro, df_pseudo_obs)\n",
    "defaults = [cropd[\"TDWI\"], cropd[\"SPAN\"]]\n",
    "error = objfunc_calculator(defaults)\n",
    "print(\"Objective function value with default parameters (%s): %s\" % (defaults, error))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's try to find the optimum by brute force\n",
    "We will now find the optimum by simply trying combinations of TDWI/SPAN across a grid in certain step sizes. Of course in reality we do not know where the true value is and you may need to make the step size smaller. Note that with small step sizes (or more parameters), the number of function calls increases exponentially and this approach becomes impractical.\n",
    "\n",
    "With the current example, we do not reach an error of zero because with our default stepsize (0.05) TWDI does not hit the true value (0.51)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Progress: |██████████████████████████████████████████████████| 100.0% Complete\n",
      " Progress: |██████████████████████████████████████████████████| 101.0% Complete\n",
      "optimum at TDWI: 0.5000000000000001, SPAN: 37\n",
      "minimum value =  0.012007708819001183\n",
      "With 100 function calls\n"
     ]
    }
   ],
   "source": [
    "lowest_error = 99999\n",
    "best_values = None\n",
    "\n",
    "# min/max and values for TDWI\n",
    "TDWI_range = [0.1, 0.6]\n",
    "stepsize1 = 0.05\n",
    "TDWI_values = np.arange(TDWI_range[0], TDWI_range[1], stepsize1)\n",
    "\n",
    "# min/max and values for SPAN\n",
    "SPAN_range = [30, 40]\n",
    "stepsize2 = 1\n",
    "SPAN_values = np.arange(SPAN_range[0], SPAN_range[1], stepsize2)\n",
    "\n",
    "nruns = len(TDWI_values) * len(SPAN_values)\n",
    "printProgressBar(0, nruns, prefix = 'Progress:', suffix = 'Complete', length = 50)\n",
    "ncalls = 0\n",
    "for p1, p2 in product(TDWI_values, SPAN_values):\n",
    "    ncalls += 1\n",
    "    error = objfunc_calculator([p1, p2])\n",
    "    if error <  lowest_error:\n",
    "        best_values = [p1, p2]\n",
    "        lowest_error = error\n",
    "    printProgressBar(ncalls+1, nruns, prefix = ' Progress:', suffix = 'Complete', length = 50)\n",
    "\n",
    "print(\"\\noptimum at TDWI: %s, SPAN: %s\" % (best_values[0], best_values[1]))\n",
    "print(\"minimum value = \",  lowest_error)\n",
    "print(\"With %i function calls\" % ncalls)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Try to do smart optimization: use NLOPT\n",
    "The [NLOPT library](https://nlopt.readthedocs.io/en/latest/) provides many algorithms for numerical optimization. We use NLOPT a lot for optimization of crop simulation models but only the global search algorithms (the 'GN' algorithms) and the local search algorithms that do not require analytical gradients (the 'LN' algorithms) because computing analytical gradients on crop simulation models is difficult. My experience is that the NLOPT library provides better optimization algorithms compared to [`scipy.optimize`](https://docs.scipy.org/doc/scipy/reference/optimize.html). \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".................................................\n",
      "optimum at TDWI: 0.4016699218750006, SPAN: 37.01171875\n",
      "minimum value =  0.14530121318162476\n",
      "result code =  3\n",
      "With 49 function calls\n"
     ]
    }
   ],
   "source": [
    "objfunc_calculator = ObjectiveFunctionCalculator(params, wdp, agro, df_pseudo_obs)\n",
    "# Start optimizer with the SUBPLEX algorithm for two parameters\n",
    "opt = nlopt.opt(nlopt.LN_SBPLX, 2)\n",
    "# Assign the objective function calculator\n",
    "opt.set_min_objective(objfunc_calculator)\n",
    "# lower bounds of parameters values\n",
    "opt.set_lower_bounds([TDWI_range[0], SPAN_range[0]])\n",
    "# upper bounds of parameters values\n",
    "opt.set_upper_bounds([TDWI_range[1], SPAN_range[1]])\n",
    "# the initial step size to compute numerical gradients\n",
    "opt.set_initial_step([0.02, 2.])\n",
    "# Maximum number of evaluations allowed\n",
    "opt.set_maxeval(200)\n",
    "# Relative tolerance for convergence\n",
    "opt.set_ftol_rel(0.1)\n",
    "\n",
    "# Start the optimization with the first guess\n",
    "firstguess = [0.2, 31.]\n",
    "x = opt.optimize(firstguess)\n",
    "print(\"\\noptimum at TDWI: %s, SPAN: %s\" % (x[0], x[1]))\n",
    "print(\"minimum value = \",  opt.last_optimum_value())\n",
    "print(\"result code = \", opt.last_optimize_result())\n",
    "print(\"With %i function calls\" % objfunc_calculator.n_calls)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As you can see, the NLOPT algorithm can find a solution with a similar accuracy with much less function calls (This depends on the value of `set_ftol_rel` though: lower tolerance value will require more function evaluations.\n",
    "\n",
    "## Plot pseudo observations and simulated LAI after optimization \n",
    "The figure below shows the pseudo observations and the final results after the model parameters were optimized given the observations. As you can see: they fit quite well with the pseudo observations and are a large improvement over the \"first guess\" "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      ".."
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# rerun with the best parameters found\n",
    "error = objfunc_calculator(x)\n",
    "fig, axes = plt.subplots(figsize=(12,8))\n",
    "axes.plot_date(df_pseudo_obs.index, df_pseudo_obs.LAI, label=\"Pseudo observation\")\n",
    "axes.plot_date(objfunc_calculator.df_simulations.index, objfunc_calculator.df_simulations.LAI, \"k:\", label=\"optimized\")\n",
    "# rerun to show the first guess for the first guess\n",
    "error = objfunc_calculator(firstguess)\n",
    "axes.plot_date(objfunc_calculator.df_simulations.index, objfunc_calculator.df_simulations.LAI, \"g:\", label=\"first guess\")\n",
    "axes.set_title(\"Pseudo LAI observations with optimized model.\")\n",
    "r = fig.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}