Load support

.gitlab-ci.yml

@@ -7,13 +7,14 @@ stages:
 
 # ===== TEST PyWake debian=====
 test_PyWake:  # name the job what we like
+  image: continuumio/anaconda3
   stage:  # build, test, deploy defined by default [2]
     test
   script:
-  - pip install -e .
-  - py.test
+  - pip install -e .[test]
+  - pytest
   tags:  # only runners with this tag can do the job [3]
-  - python
+  - ci-ubuntu
 
   
 # ===== Check code style =====
@@ -31,7 +32,8 @@ test_PyWake_windows:  # name the job what we like
   stage:  # build, test, deploy defined by default [2]
     test
   script:  # runs on windows machine due to tag below
-  - source activate pyTopfarm
+  - source activate py38
+  - pip install -e .[test]
   - pytest --cov-report term-missing:skip-covered --cov=py_wake --cov-config .coveragerc
   tags:  # tag for shared runner on windows machine
   - ANMH_old

docs/notebooks/ChangeLog.ipynb

@@ -11,7 +11,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "## PyWake 2.xxx (March, 2021)\n",
+    "## PyWake 2.2 (March 26, 2021)\n",
     "### New features and API changes\n",
     "- All DeficitModels should inherrit either `WakeDeficitModel` or `BlockageDeficitModel`\n",
     "- All Sites are now subclasses of XRSite\n",
@@ -26,6 +26,10 @@
     "    - `PowerCtNDTabular`\n",
     "    - `PowerCtXr`\n",
     "    - `CubePowerSimpleCt`\n",
+    "- Support for time series of wd and ws, see [here](https://topfarm.pages.windenergy.dtu.dk/PyWake/notebooks/RunWindFarmSimulation.html#Time-series). Possible use cases:\n",
+    "    - Time-dependent inflow, e.g. measurements of wd, ws, ti, shear, density, etc.\n",
+    "    - Time-dependent operation, e.g. periods of failure or maintaince of a wind turbine\n",
+    "- Added support for load surrogates to predict wind turbine loads\n",
     "\n",
     "### New models\n",
     "- BlockageDeficitModels (see [here](https://topfarm.pages.windenergy.dtu.dk/PyWake/notebooks/EngineeringWindFarmModels.html#Blockage-deficit-models)\n",
@@ -127,7 +131,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.9.2"
+   "version": "3.8.8"
   },
   "toc": {
    "base_numbering": 1,

py_wake/examples/data/dtu10mw/__init__.py

+from py_wake.examples.data.dtu10mw._dtu10mw import DTU10MW, DTU10WM_RWT, ct_curve, power_curve

py_wake/examples/data/dtu10mw.py → py_wake/examples/data/dtu10mw/_dtu10mw.py

@@ -23,7 +23,7 @@ power_curve = np.array([[4, 280.2],
                         [23, 10639.5],
                         [24, 10643.6],
                         [25, 10635.7],
-                        ]) * [1, .94]
+                        ]) * [1, .94]  # 6% loss
 ct_curve = np.array([[4, 0.923],
                      [5, 0.919],
                      [6, 0.904],

py_wake/examples/data/hornsrev1.py

@@ -94,14 +94,14 @@ HornsrevV80 = V80
 
 
 class Hornsrev1Site(UniformWeibullSite):
-    def __init__(self, shear=None):
+    def __init__(self, ti=.1, shear=None):
         f = [3.597152, 3.948682, 5.167395, 7.000154, 8.364547, 6.43485,
              8.643194, 11.77051, 15.15757, 14.73792, 10.01205, 5.165975]
         a = [9.176929, 9.782334, 9.531809, 9.909545, 10.04269, 9.593921,
              9.584007, 10.51499, 11.39895, 11.68746, 11.63732, 10.08803]
         k = [2.392578, 2.447266, 2.412109, 2.591797, 2.755859, 2.595703,
              2.583984, 2.548828, 2.470703, 2.607422, 2.626953, 2.326172]
-        UniformWeibullSite.__init__(self, np.array(f) / np.sum(f), a, k, .1, shear=shear)
+        UniformWeibullSite.__init__(self, np.array(f) / np.sum(f), a, k, ti=ti, shear=shear)
         self.initial_position = np.array([wt_x, wt_y]).T
 
 

py_wake/examples/data/iea34_130rwt/__init__.py

+from ._iea34_130rwt import IEA34_130_1WT_Surrogate, IEA34_130_2WT_Surrogate

py_wake/examples/data/iea34_130rwt/_iea34_130rwt.py

+from py_wake.wind_turbines import WindTurbine
+import numpy as np
+from pathlib import Path
+from py_wake.utils.tensorflow_surrogate_utils import TensorflowSurrogate
+import inspect
+from py_wake.wind_turbines.power_ct_functions import PowerCtSurrogate
+from py_wake.wind_turbines.wind_turbine_functions import FunctionSurrogates
+from py_wake.examples.data import example_data_path
+
+
+class IEA34_130_PowerCtSurrogate(PowerCtSurrogate):
+    def __init__(self, name, input_parser):
+        surrogate_path = Path(__file__).parent / name
+        PowerCtSurrogate.__init__(
+            self,
+            power_surrogate=TensorflowSurrogate(surrogate_path / "electrical_power", 'operating'),
+            power_unit='W',
+            ct_surrogate=TensorflowSurrogate(surrogate_path / 'thrust', 'operating'),
+            input_parser=input_parser)
+
+        ws_idx = self.function_surrogate_lst[0].input_channel_names.index('ws')
+        self.ws_cutin = self.function_surrogate_lst[0].input_scaler.data_min_[ws_idx]  # .wind_speed_cut_in
+        self.ws_cutout = self.function_surrogate_lst[0].input_scaler.data_max_[ws_idx]  # .wind_speed_cut_out
+        ti_key = [k for k in list(inspect.signature(input_parser).parameters) if k[:2] == 'TI'][0]
+        thrust_idle = PowerCtSurrogate._power_ct(self, np.array([self.ws_cutout]), **{ti_key: .1})[1] * 1000
+        self.ct_idle = thrust_idle / (1 / 2 * 1.225 * (65**2 * np.pi) * self.ws_cutout**2)
+
+    def _power_ct(self, ws, **kwargs):
+        power, thrust = PowerCtSurrogate._power_ct(self, ws, **kwargs)
+        ct = thrust * 1000 / (1 / 2 * 1.225 * (65**2 * np.pi) * ws**2)
+        power[(ws < self.ws_cutin) | (ws > self.ws_cutout)] = 0
+        ct[(ws < self.ws_cutin) | (ws > self.ws_cutout)] = self.ct_idle
+        return power, ct
+
+
+class TreeRegionLoadSurrogates(FunctionSurrogates):
+    def __init__(self, function_surrogate_lst, input_parser):
+        FunctionSurrogates.__init__(self, function_surrogate_lst, input_parser)
+        self.ws_cutin = function_surrogate_lst[0][0].wind_speed_cut_in
+        self.ws_cutout = function_surrogate_lst[0][0].wind_speed_cut_out
+
+    def __call__(self, ws, run_only=slice(None), **kwargs):
+        ws_flat = ws.ravel()
+        x = self.get_input(ws=ws, **kwargs)
+        x = np.array([self.fix_shape(v, ws).ravel() for v in x]).T
+
+        def predict(fs):
+            output = np.empty(len(x))
+            for fs_, m in zip(fs, [ws_flat < self.ws_cutin,
+                                   (self.ws_cutin <= ws_flat) & (ws_flat <= self.ws_cutout),
+                                   ws_flat > self.ws_cutout]):
+                if m.sum():
+                    output[m] = fs_.predict_output(x[m], bounds='ignore')[:, 0]
+            return output
+        return [predict(fs).reshape(ws.shape) for fs in np.asarray(self.function_surrogate_lst)[run_only]]
+
+    @property
+    def output_keys(self):
+        return [fs[0].output_channel_name for fs in self.function_surrogate_lst]
+
+    @property
+    def wohler_exponents(self):
+        return [fs[0].wohler_exponent for fs in self.function_surrogate_lst]
+
+
+class IEA34_130_Base(WindTurbine):
+    def __init__(self, powerCtFunction, loadFunction):
+        WindTurbine.__init__(self, 'IEA 3.4MW', diameter=130, hub_height=110,
+                             powerCtFunction=powerCtFunction,
+                             loadFunction=loadFunction)
+
+
+class IEA34_130_1WT_Surrogate(IEA34_130_Base):
+
+    def __init__(self):
+        sensors = ['del_blade_flap', 'del_blade_edge', 'del_tower_bottom_fa', 'del_tower_bottom_ss',
+                   'del_tower_top_torsion']
+        surrogate_path = Path(example_data_path) / 'iea34_130rwt' / 'one_turbine'
+        set_names = ['below_cut_in', 'operating', 'above_cut_out']
+        loadFunction = TreeRegionLoadSurrogates(
+            [[TensorflowSurrogate(surrogate_path / s, n) for n in set_names] for s in sensors],
+            input_parser=lambda ws, TI_eff=.1, Alpha=0: [TI_eff, ws, Alpha])
+        powerCtFunction = IEA34_130_PowerCtSurrogate(
+            'one_turbine',
+            input_parser=lambda ws, TI_eff, Alpha=0: [TI_eff, ws, Alpha])
+        IEA34_130_Base.__init__(self, powerCtFunction=powerCtFunction, loadFunction=loadFunction)
+
+
+class IEA34_130_2WT_Surrogate(IEA34_130_Base):
+    def __init__(self):
+        sensors = ['del_blade_flap']
+        surrogate_path = Path(example_data_path) / 'iea34_130rwt' / 'two_turbines'
+        set_names = ['below_cut_in', 'operating', 'above_cut_out']
+        loadFunction = TreeRegionLoadSurrogates(
+            [[TensorflowSurrogate(surrogate_path / s, n) for n in set_names] for s in sensors],
+            input_parser=self.get_input)
+        self.max_dist = loadFunction.function_surrogate_lst[0][0].input_scaler.data_max_[4]
+        self.max_angle = loadFunction.function_surrogate_lst[0][0].input_scaler.data_max_[3]
+
+        powerCtFunction = IEA34_130_PowerCtSurrogate(
+            'two_turbines',
+            input_parser=(lambda ws, TI=.1, Alpha=0, get_input=self.get_input:
+                          get_input(ws, TI=TI, Alpha=Alpha, dw_ijl=np.array([0]), hcw_ijl=0)))
+        IEA34_130_Base.__init__(self, powerCtFunction=powerCtFunction, loadFunction=loadFunction)
+
+    def get_input(self, ws, dw_ijl, hcw_ijl, TI, Alpha=0):
+        # ['ws','ti', 'shear', 'wdir', 'dist']
+        dist = np.atleast_1d((np.hypot(dw_ijl, hcw_ijl) / 130))
+        wd = np.atleast_1d(np.rad2deg(np.arctan2(hcw_ijl, dw_ijl)))
+        unwaked = (dist == 0) | (dist > self.max_dist) | (np.abs(wd) > self.max_angle)
+        dist[unwaked] = 20
+        wd[unwaked] = 20
+        return [ws, TI, Alpha, wd, dist]
+
+
+def main():
+    if __name__ == '__main__':
+        from py_wake.examples.data.hornsrev1 import Hornsrev1Site
+        from py_wake.turbulence_models.stf import STF2017TurbulenceModel
+        from py_wake import NOJ
+        import matplotlib.pyplot as plt
+
+        u = np.arange(3, 28, .1)
+        # ===============================================================================================================
+        # IEA34_130_1WT_Surrogate
+        # ===============================================================================================================
+        wt = IEA34_130_1WT_Surrogate()
+        # plot power/ct curves
+        ax1 = plt.gca()
+        ax2 = plt.twinx()
+        for ti in [0.01, .05, .1, .3]:
+            power, ct = wt.power_ct(u, TI_eff=ti)
+            ax1.plot(u, power / 1000, label=f'TI={ti}')
+            ax2.plot(u, ct, '--')
+        ax1.legend()
+        ax1.set_ylabel('Power [kW]')
+        ax2.set_ylabel('Ct')
+
+        plt.figure()
+        ax1 = plt.gca()
+        ax2 = plt.twinx()
+        for alpha in [-0.09, .1, .3, .49]:
+            power, ct = wt.power_ct(u, TI_eff=.1, Alpha=alpha)
+            ax1.plot(u, power / 1000, label=f'Alpha={alpha}')
+            ax2.plot(u, ct, '--')
+        ax1.legend()
+        ax1.set_ylabel('Power [kW]')
+        ax2.set_ylabel('Ct')
+
+        # plot load curves
+        sensors = wt.loadFunction.output_keys
+        axes = [plt.figure().gca() for _ in sensors]
+        for ti in [0.01, .05, .1, .3]:
+            loads = wt.loads(u, TI_eff=ti)
+            for ax, l in zip(axes, loads):
+                ax.plot(u, l, label=f'TI={ti}')
+        for alpha in [-0.09, .1, .3, .49]:
+            loads = wt.loads(u, TI_eff=.1, Alpha=alpha)
+            for ax, l in zip(axes, loads):
+                ax.plot(u, l, '--', label=f'Alpha={alpha}')
+        for ax, s in zip(axes, sensors):
+            ax.set_title(s)
+            ax.legend()
+
+        # plot loads as function of wd and ws
+        plt.figure()
+        site = Hornsrev1Site()
+        x, y = [0, 1000], [0, 0]
+        sim_res = NOJ(site, wt, turbulenceModel=STF2017TurbulenceModel())(x, y, ws=np.arange(6, 25), Alpha=.12)
+        load_wd_averaged = sim_res.loads(normalize_probabilities=True, method='OneWT_WDAvg')
+        loads = sim_res.loads(normalize_probabilities=True, method='OneWT')
+        loads.DEL.isel(sensor=0, wt=0).plot()
+
+        for s in load_wd_averaged.sensor:
+            print(s.item(), load_wd_averaged.LDEL.sel(sensor=s, wt=0).item(), loads.LDEL.sel(sensor=s, wt=0).item())
+        plt.show()
+
+        # =======================================================================================================================
+        # IEA34_130_2WTSurrogate
+        # =======================================================================================================================
+        wt = IEA34_130_2WT_Surrogate()
+        # plot power/ct curves
+        plt.figure()
+        ax1 = plt.gca()
+        ax2 = plt.twinx()
+        for ti in [0.01, .05, .1, .3]:
+            power, ct = wt.power_ct(u, TI=ti)
+            ax1.plot(u, power, label=f'TI={ti}')
+            ax2.plot(u, ct, '--')
+        ax1.legend()
+        ax1.set_ylabel('Power [kW]')
+        ax2.set_ylabel('Ct')
+
+        plt.figure()
+        ax1 = plt.gca()
+        ax2 = plt.twinx()
+        for alpha in [-0.09, .1, .3, .49]:
+            power, ct = wt.power_ct(u, TI=.1, Alpha=alpha)
+            ax1.plot(u, power / 1000, label=f'Alpha={alpha}')
+            ax2.plot(u, ct, '--')
+        ax1.set_ylabel('Power [kW]')
+        ax2.set_ylabel('Ct')
+        ax1.legend()
+
+        # plot load curves
+        sensors = wt.loadFunction.output_keys
+        axes = [plt.figure().gca() for _ in sensors]
+        for ti in [0.01, .05, .1, .3]:
+            loads = wt.loads(u, TI=ti, Alpha=.12, dw_ijl=0, hcw_ijl=0)
+            for ax, l in zip(axes, loads):
+                ax.plot(u, l, label=f'TI={ti}')
+        for alpha in [-0.09, .1, .3, .49]:
+            loads = wt.loads(u, TI=.1, Alpha=alpha, dw_ijl=np.array([1000]), hcw_ijl=0)
+            for ax, l in zip(axes, loads):
+                ax.plot(u, l, '--', label=f'Alpha={alpha}')
+        for ax, s in zip(axes, sensors):
+            ax.set_title(s)
+            ax.legend()
+
+        # plot loads as function of wd and ws
+        plt.figure()
+        site = Hornsrev1Site()
+        x, y = [0, 1000], [0, 0]
+        sim_res = NOJ(site, wt, turbulenceModel=STF2017TurbulenceModel())(x, y, ws=np.arange(3, 28), Alpha=.12)
+        loads = sim_res.loads(normalize_probabilities=True, method='TwoWT')
+        loads.DEL.isel(sensor=0, wt=0).plot()
+
+        for s in loads.sensor:
+            print(s.item(), loads.LDEL.sel(sensor=s, wt=0).item())
+        plt.show()
+
+
+main()

py_wake/examples/data/iea34_130rwt/one_turbine/del_blade_edge/extra_data.json

+{
+    "input_channel_names": [
+        "ti",
+        "ws",
+        "shear"
+    ],
+    "output_channel_name": "MomentMy Mbdy:blade1 nodenr:   1 coo: blade1  blade root moment blade1",
+    "wind_speed_cut_in": 4.0,
+    "wind_speed_cut_out": 25.0,
+    "wohler_exponent": 10,
+    "input_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 290,
+            "scale_": [
+                4.093931847815023,
+                1.5527559984108794,
+                10.034296914116865
+            ],
+            "min_": [
+                -3.00460564098673,
+                -3.206397077344649,
+                -2.0024497794364287
+            ],
+            "data_min_": [
+                0.0011249921,
+                0.1329230591,
+                -0.0994140625
+            ],
+            "data_max_": [
+                1.4667087446,
+                3.997020191,
+                0.4985351562
+            ],
+            "data_range_": [
+                1.4655837525,
+                3.8640971319,
+                0.5979492187
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 138,
+            "scale_": [
+                17.108732300045514,
+                1.2235952583871186,
+                10.058939096267192
+            ],
+            "min_": [
+                -3.028362517654116,
+                -33.59896481247119,
+                -2.0029469543104126
+            ],
+            "data_min_": [
+                0.0016577802,
+                25.0074234946,
+                -0.0991210938
+            ],
+            "data_max_": [
+                0.3523558854,
+                29.9110057526,
+                0.4973632812
+            ],
+            "data_range_": [
+                0.3506981052,
+                4.903582258,
+                0.596484375
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                11.79559835553993,
+                0.28710421679370607,
+                10.014669925814587
+            ],
+            "min_": [
+                -3.000817363012889,
+                -4.148878985987874,
+                -2.0014669922476362
+            ],
+            "data_min_": [
+                6.92939e-05,
+                4.0016095856,
+                -0.0997070313
+            ],
+            "data_max_": [
+                0.5087336125,
+                24.8999442287,
+                0.4994140625
+            ],
+            "data_range_": [
+                0.5086643186,
+                20.8983346431,
+                0.5991210938
+            ]
+        }
+    },
+    "output_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 290,
+            "scale_": [
+                0.003217829437758253
+            ],
+            "min_": [
+                -2.241711711671581
+            ],
+            "data_min_": [
+                385.8848754073917
+            ],
+            "data_max_": [
+                1007.4218582355832
+            ],
+            "data_range_": [
+                621.5369828281914
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 138,
+            "scale_": [
+                0.006414867921372538
+            ],
+            "min_": [
+                -5.543090761343728
+            ],
+            "data_min_": [
+                708.2126735933917
+            ],
+            "data_max_": [
+                1019.9883834777
+            ],
+            "data_range_": [
+                311.7757098843083
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                0.0021491953622476507
+            ],
+            "min_": [
+                -11.971056356553449
+            ],
+            "data_min_": [
+                5104.727354836558
+            ],
+            "data_max_": [
+                6035.308182960242
+            ],
+            "data_range_": [
+                930.5808281236841
+            ]
+        }
+    }
+}
\ No newline at end of file

py_wake/examples/data/iea34_130rwt/one_turbine/del_blade_flap/extra_data.json

+{
+    "input_channel_names": [
+        "ti",
+        "ws",
+        "shear"
+    ],
+    "output_channel_name": "MomentMx Mbdy:blade1 nodenr:   1 coo: blade1  blade root moment blade1",
+    "wind_speed_cut_in": 4.0,
+    "wind_speed_cut_out": 25.0,
+    "wohler_exponent": 10,
+    "input_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 290,
+            "scale_": [
+                4.093931847815023,
+                1.5527559984108794,
+                10.034296914116865
+            ],
+            "min_": [
+                -3.00460564098673,
+                -3.206397077344649,
+                -2.0024497794364287
+            ],
+            "data_min_": [
+                0.0011249921,
+                0.1329230591,
+                -0.0994140625
+            ],
+            "data_max_": [
+                1.4667087446,
+                3.997020191,
+                0.4985351562
+            ],
+            "data_range_": [
+                1.4655837525,
+                3.8640971319,
+                0.5979492187
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 138,
+            "scale_": [
+                17.108732300045514,
+                1.2235952583871186,
+                10.058939096267192
+            ],
+            "min_": [
+                -3.028362517654116,
+                -33.59896481247119,
+                -2.0029469543104126
+            ],
+            "data_min_": [
+                0.0016577802,
+                25.0074234946,
+                -0.0991210938
+            ],
+            "data_max_": [
+                0.3523558854,
+                29.9110057526,
+                0.4973632812
+            ],
+            "data_range_": [
+                0.3506981052,
+                4.903582258,
+                0.596484375
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                11.79559835553993,
+                0.28710421679370607,
+                10.014669925814587
+            ],
+            "min_": [
+                -3.000817363012889,
+                -4.148878985987874,
+                -2.0014669922476362
+            ],
+            "data_min_": [
+                6.92939e-05,
+                4.0016095856,
+                -0.0997070313
+            ],
+            "data_max_": [
+                0.5087336125,
+                24.8999442287,
+                0.4994140625
+            ],
+            "data_range_": [
+                0.5086643186,
+                20.8983346431,
+                0.5991210938
+            ]
+        }
+    },
+    "output_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 290,
+            "scale_": [
+                0.0044128736952490375
+            ],
+            "min_": [
+                -16.790881035417325
+            ],
+            "data_min_": [
+                3578.366870644408
+            ],
+            "data_max_": [
+                4031.586277796992
+            ],
+            "data_range_": [
+                453.2194071525837
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 138,
+            "scale_": [
+                0.004608273201192155
+            ],
+            "min_": [
+                -15.907513088691113
+            ],
+            "data_min_": [
+                3234.945594118542
+            ],
+            "data_max_": [
+                3668.947640586317
+            ],
+            "data_range_": [
+                434.002046467775
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                0.00032756328073293837
+            ],
+            "min_": [
+                -1.359284416662768
+            ],
+            "data_min_": [
+                1096.839718599875
+            ],
+            "data_max_": [
+                7202.530184041866
+            ],
+            "data_range_": [
+                6105.690465441991
+            ]
+        }
+    }
+}
\ No newline at end of file

py_wake/examples/data/iea34_130rwt/one_turbine/del_tower_bottom_fa/extra_data.json

+{
+    "input_channel_names": [
+        "ti",
+        "ws",
+        "shear"
+    ],
+    "output_channel_name": "MomentMx Mbdy:tower nodenr:   1 coo: tower  tower bottom moment",
+    "wind_speed_cut_in": 4.0,
+    "wind_speed_cut_out": 25.0,
+    "wohler_exponent": 4,
+    "input_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 290,
+            "scale_": [
+                4.093931847815023,
+                1.5527559984108794,
+                10.034296914116865
+            ],
+            "min_": [
+                -3.00460564098673,
+                -3.206397077344649,
+                -2.0024497794364287
+            ],
+            "data_min_": [
+                0.0011249921,
+                0.1329230591,
+                -0.0994140625
+            ],
+            "data_max_": [
+                1.4667087446,
+                3.997020191,
+                0.4985351562
+            ],
+            "data_range_": [
+                1.4655837525,
+                3.8640971319,
+                0.5979492187
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 138,
+            "scale_": [
+                17.108732300045514,
+                1.2235952583871186,
+                10.058939096267192
+            ],
+            "min_": [
+                -3.028362517654116,
+                -33.59896481247119,
+                -2.0029469543104126
+            ],
+            "data_min_": [
+                0.0016577802,
+                25.0074234946,
+                -0.0991210938
+            ],
+            "data_max_": [
+                0.3523558854,
+                29.9110057526,
+                0.4973632812
+            ],
+            "data_range_": [
+                0.3506981052,
+                4.903582258,
+                0.596484375
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -3.0,
+                3.0
+            ],
+            "copy": true,
+            "n_features_in_": 3,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                11.79559835553993,
+                0.28710421679370607,
+                10.014669925814587
+            ],
+            "min_": [
+                -3.000817363012889,
+                -4.148878985987874,
+                -2.0014669922476362
+            ],
+            "data_min_": [
+                6.92939e-05,
+                4.0016095856,
+                -0.0997070313
+            ],
+            "data_max_": [
+                0.5087336125,
+                24.8999442287,
+                0.4994140625
+            ],
+            "data_range_": [
+                0.5086643186,
+                20.8983346431,
+                0.5991210938
+            ]
+        }
+    },
+    "output_scalers": {
+        "below_cut_in": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 290,
+            "scale_": [
+                0.0006468022347309646
+            ],
+            "min_": [
+                -2.4917984877225337
+            ],
+            "data_min_": [
+                2306.4213566656003
+            ],
+            "data_max_": [
+                5398.556622450348
+            ],
+            "data_range_": [
+                3092.135265784748
+            ]
+        },
+        "above_cut_out": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 138,
+            "scale_": [
+                0.0005535856485690355
+            ],
+            "min_": [
+                -3.5032795237167074
+            ],
+            "data_min_": [
+                4521.937174830017
+            ],
+            "data_max_": [
+                8134.747595710333
+            ],
+            "data_range_": [
+                3612.810420880316
+            ]
+        },
+        "operating": {
+            "feature_range": [
+                -1.0,
+                1.0
+            ],
+            "copy": true,
+            "n_features_in_": 1,
+            "n_samples_seen_": 1572,
+            "scale_": [
+                0.0001684019418641057
+            ],
+            "min_": [
+                -2.0006685073252397
+            ],
+            "data_min_": [
+                5942.1435183493495
+            ],
+            "data_max_": [
+                17818.491129672784
+            ],
+            "data_range_": [
+                11876.347611323436
+            ]
+        }
+    }
+}
\ No newline at end of file