def test_pvengine_float_inputs_perez(params):
"""Test that PV engine works for float inputs"""
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_illum_pvrow'],
DNI)
# Run timestep
pvarray = eng.run_full_mode_timestep(0)
# Checks
assert isinstance(pvarray, OrderedPVArray)
np.testing.assert_almost_equal(
pvarray.pvrows[0].front.get_param_weighted('qinc'), 1110.1164773159298)
np.testing.assert_almost_equal(
pvarray.pvrows[1].front.get_param_weighted('qinc'), 1110.595903991)
np.testing.assert_almost_equal(
pvarray.pvrows[2].front.get_param_weighted('qinc'), 1112.37717553)
np.testing.assert_almost_equal(
pvarray.pvrows[1].back.get_param_weighted('qinc'), 116.49050349491208)
def test_pvengine_ts_inputs_perez(params_serial, df_inputs_serial_calculation,
fn_report_example):
"""Test that PV engine works for timeseries inputs"""
# Break up inputs
(timestamps, surface_tilt, surface_azimuth, solar_zenith, solar_azimuth,
dni, dhi) = breakup_df_inputs(df_inputs_serial_calculation)
albedo = params_serial['rho_ground']
# Create engine
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(
params_serial, param_names=irradiance_model.params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Fit engine
eng.fit(timestamps, dni, dhi, solar_zenith, solar_azimuth, surface_tilt,
surface_azimuth, albedo)
# Run all timesteps
report = eng.run_full_mode(fn_build_report=fn_report_example)
# Check values
np.testing.assert_array_almost_equal(report['qinc_front'],
[1066.272392, 1065.979824])
np.testing.assert_array_almost_equal(report['qinc_back'],
[135.897106, 136.01297])
np.testing.assert_array_almost_equal(report['iso_front'],
[42.816637, 42.780206])
np.testing.assert_array_almost_equal(report['iso_back'],
[1.727308, 1.726535])
np.testing.assert_array_almost_equal(report['qabs_back'],
report['qinc_back'] * 0.97)
def test_fast_pvengine_float_inputs_perez(params):
"""Test that PV engine works for float inputs"""
# Prepare some engine inputs
irradiance_model = HybridPerezOrdered()
fast_mode_pvrow_index = 1
# Create engine object
eng = PVEngine(params,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_pvrow'], DNI)
# Run timestep
pvarray = eng.run_timestep(0)
# Checks
assert isinstance(pvarray, OrderedPVArray)
np.testing.assert_almost_equal(
pvarray.pvrows[1].back.get_param_weighted('qinc'), 123.7087347744459)
def test_pvengine_float_inputs_iso(params):
"""Test that PV engine works for float inputs"""
irradiance_model = IsotropicOrdered()
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_illum_pvrow'],
DNI)
# Run timestep
pvarray = eng.run_full_mode(fn_build_report=lambda pvarray: pvarray)
# Checks
assert isinstance(pvarray, OrderedPVArray)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[0].front.get_param_weighted('qinc'), 1099.22245374)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[1].front.get_param_weighted('qinc'), 1099.6948573)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[2].front.get_param_weighted('qinc'), 1102.76149246)
# Check absorbed
np.testing.assert_almost_equal(
pvarray.ts_pvrows[1].front.get_param_weighted('qabs'),
1099.6948573 * 0.99)
def test_fast_mode_8760(params, df_inputs_clearsky_8760):
"""Test fast mode with 1 PV row to make sure that is consistent
after the vectorization update: should get exact same values
"""
# Get MET data
df_inputs = df_inputs_clearsky_8760
timestamps = df_inputs.index
dni = df_inputs.dni.values
dhi = df_inputs.dhi.values
solar_zenith = df_inputs.solar_zenith.values
solar_azimuth = df_inputs.solar_azimuth.values
surface_tilt = df_inputs.surface_tilt.values
surface_azimuth = df_inputs.surface_azimuth.values
# Run simulation for only 1 PV row
params.update({'n_pvrows': 1})
# Run engine
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray)
eng.fit(timestamps, dni, dhi, solar_zenith, solar_azimuth, surface_tilt,
surface_azimuth, params['rho_ground'])
qinc = eng.run_fast_mode(
fn_build_report=lambda pvarray:
(pvarray.ts_pvrows[0].back.get_param_weighted('qinc')),
pvrow_index=0)
# Check than annual energy on back is consistent
np.testing.assert_allclose(np.nansum(qinc) / 1e3, 342.848005)
def test_fast_mode_loop_like(params):
"""Test value of older and decomissioned loop like fast mode"""
# Prepare some engine inputs
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(
params, param_names=irradiance_model.params)
fast_mode_pvrow_index = 1
# Create engine object
eng = PVEngine(pvarray,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_illum_pvrow'],
DNI)
# Run timestep
pvarray = _fast_mode_with_loop(eng.pvarray, eng.irradiance,
eng.vf_calculator, fast_mode_pvrow_index, 0)
# Checks
assert isinstance(pvarray, OrderedPVArray)
np.testing.assert_almost_equal(
pvarray.pvrows[1].back.get_param_weighted('qinc'), 119.0505580124769)
def test_vf_matrix_subset_calculation(params):
"""Check that the vf matrix subset is calculated correctly"""
# Run in fast mode
irradiance_model = HybridPerezOrdered()
fast_mode_pvrow_index = 1
eng = PVEngine(params,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index)
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
pvarray_fast = eng.run_timestep(0)
vf_mat_fast = pvarray_fast.vf_matrix
# Run in full mode
irradiance_model = HybridPerezOrdered()
eng = PVEngine(params, irradiance_model=irradiance_model)
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
pvarray_full = eng.run_timestep(0)
vf_mat_full = pvarray_full.vf_matrix
index_of_back_side_surface = 13
np.testing.assert_array_almost_equal(
vf_mat_fast, vf_mat_full[[index_of_back_side_surface], :])
def test_check_direct_shading_continuity():
"""Make sure the calculation is correct when direct shading happens.
- before v1.3.0, there's a discontinuity (big jump) in prediction
when direct shading happens. The values are the same as >=v1.3.0 for
no direct shading, but they are different with direct shading.
- starting at v1.3.0, the values are more continuous (no big change) when
going from no direct shading to direct shading, which means it's
most certainly a better implementation.
The issue before v1.3.0 could be due to the fact that shadows are merged,
and there might be a piece of geometry lost there, but not entirely sure.
Since it's still relatively small, will not dig further.
Here, we're testing the outputs at 2 timestamps, right before and
right after direct shading, when varying only solar zenith.
"""
# Prepare inputs
n = 2
inputs = {
'solar_zenith': [81.275, 81.276], # right at the limit of direct shadg
'solar_azimuth': [295.9557133] * n,
'surface_tilt': [15.18714669] * n,
'surface_azimuth': [270.] * n,
'dni': [1000.] * n,
'dhi': [100.] * n,
'albedo': [0.2] * n,
'times': [dt.datetime(2014, 6, 25, 3)] * n}
# Array parameters
params = {'n_pvrows': 3,
'axis_azimuth': 0.0,
'pvrow_height': 1.5,
'pvrow_width': 2.5,
'gcr': 0.4}
# Create engine and fit to inputs
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray)
eng.fit(np.array(inputs['times']),
np.array(inputs['dni']),
np.array(inputs['dhi']),
np.array(inputs['solar_zenith']),
np.array(inputs['solar_azimuth']),
np.array(inputs['surface_tilt']),
np.array(inputs['surface_azimuth']),
np.array(inputs['albedo']))
# Check there we are indeed right at the limit of direct shading
np.testing.assert_array_equal(pvarray.has_direct_shading, [False, True])
# Run simulation and get output
pvarray = eng.run_full_mode(fn_build_report=lambda pvarray: pvarray)
out = pvarray.ts_pvrows[1].back.get_param_weighted('qinc')
# Check expected outputs: before v1.3.0, expected output is
# [20.4971271991293, 21.389095477613356], which shows discontinuity
expected_out = [20.497127, 20.50229]
np.testing.assert_allclose(out, expected_out)
def run_pvfactors_simulation(data: DataFrame,
pvarray_params: dict) -> OrderedPVArray:
"Fit and run simulation"
pvarray = OrderedPVArray.init_from_dict(pvarray_params)
engine = PVEngine(pvarray)
pvarray = engine.fit(data.index, data.dni, data.dhi, data.zenith,
data.azimuth, data.surface_tilt, data.surface_azimuth,
data.albedo)
fn_build_report = lambda pvarray: pvarray
return engine.run_full_mode(fn_build_report)
def test_pvengine_float_inputs_perez_transparency_spacing_fast(params):
"""Test that module transparency and spacing are having the
expected effect to calculated PV back side irradiance"""
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# --- with 0 transparency and spacing
# Create models
irr_params = {'module_transparency': 0.,
'module_spacing_ratio': 0.}
irradiance_model = HybridPerezOrdered(**irr_params)
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Fit engine
eng.fit(timestamps, DNI, DHI,
params['solar_zenith'],
params['solar_azimuth'],
params['surface_tilt'],
params['surface_azimuth'],
params['rho_ground'])
# Run timestep
def fn_report(pvarray): return (pvarray.ts_pvrows[1]
.back.get_param_weighted('qinc'))
no_spacing_transparency_back_qinc = \
eng.run_fast_mode(fn_build_report=fn_report, pvrow_index=1)
# --- with non-0 transparency and spacing
# Create models
irr_params = {'module_transparency': 0.1,
'module_spacing_ratio': 0.1}
irradiance_model = HybridPerezOrdered(**irr_params)
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Fit engine
eng.fit(timestamps, DNI, DHI,
params['solar_zenith'],
params['solar_azimuth'],
params['surface_tilt'],
params['surface_azimuth'],
params['rho_ground'])
# Run timestep
w_spacing_transparency_back_qinc = \
eng.run_fast_mode(fn_build_report=fn_report, pvrow_index=1)
# Checks
expected_back_qinc = 134.7143531 # higher than when params are 0
np.testing.assert_almost_equal(
w_spacing_transparency_back_qinc, expected_back_qinc)
assert no_spacing_transparency_back_qinc < w_spacing_transparency_back_qinc
def test_engine_w_faoi_fn_in_irradiance_vfcalcs(params, pvmodule_canadian):
"""Run PV engine calcs with faoi functions for AOI losses"""
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
pvarray = OrderedPVArray.init_from_dict(params)
# create faoi function
faoi_fn = faoi_fn_from_pvlib_sandia(pvmodule_canadian)
# create vf_calculator with faoi function
vfcalculator = VFCalculator(faoi_fn_front=faoi_fn, faoi_fn_back=faoi_fn)
# create irradiance model with faoi function
irradiance_model = HybridPerezOrdered(faoi_fn_front=faoi_fn,
faoi_fn_back=faoi_fn)
eng = PVEngine(pvarray, irradiance_model=irradiance_model,
vf_calculator=vfcalculator)
# Make sure aoi methods are available
assert eng.vf_calculator.vf_aoi_methods is not None
# Fit engine
eng.fit(timestamps, DNI, DHI,
params['solar_zenith'],
params['solar_azimuth'],
params['surface_tilt'],
params['surface_azimuth'],
params['rho_ground'])
# Run timestep
pvarray = eng.run_full_mode(fn_build_report=lambda pvarray: pvarray)
# Checks
np.testing.assert_almost_equal(
pvarray.ts_pvrows[0].front.get_param_weighted('qinc'),
1110.1164773159298)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[1].front.get_param_weighted('qinc'), 1110.595903991)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[2].front.get_param_weighted('qinc'), 1112.37717553)
np.testing.assert_almost_equal(
pvarray.ts_pvrows[1].back.get_param_weighted('qinc'),
116.49050349491208)
# Check absorbed irradiance: calculated using faoi functions
np.testing.assert_almost_equal(
pvarray.ts_pvrows[2].front.get_param_weighted('qabs'),
[1109.1180884])
np.testing.assert_almost_equal(
pvarray.ts_pvrows[1].back.get_param_weighted('qabs'),
[114.2143503])
def test_run_fast_mode_compare_to_loop_like(params):
"""Test that PV engine works for timeseries fast mode and float inputs.
Check the left pv row back side (no obstruction, since rotation < 0)
This should be exactly the same calculated value as in loop-like fast mode
because ground is not infinite for back of left pv row"""
# Prepare some engine inputs
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(
params, param_names=irradiance_model.params)
fast_mode_pvrow_index = 1
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Create engine object
eng = PVEngine(pvarray,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index)
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Run baseline calculation
pvrow_idx = 0
time_idx = 0
pvarray_loop = _fast_mode_with_loop(eng.pvarray, eng.irradiance,
eng.vf_calculator, pvrow_idx, time_idx)
# Expected should be: 138.10421248631152
qinc_expected = pvarray_loop.pvrows[0].back.get_param_weighted('qinc')
# Run timeseries calculation
qinc = eng.run_fast_mode(
fn_build_report=lambda pvarray:
(pvarray.ts_pvrows[0].back.get_param_weighted('qinc')),
pvrow_index=0)
# Check results: value taken from loop-like fast mode
np.testing.assert_allclose(qinc, qinc_expected)
def test_run_fast_and_full_modes_sequentially(params, fn_report_example):
"""Make sure that can run fast and full modes one after the other
without making the engine crash"""
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Prepare some engine inputs
pvarray = OrderedPVArray.init_from_dict(params)
fast_mode_pvrow_index = 1
fast_mode_segment_index = 0
# Create engine object
eng = PVEngine(pvarray,
fast_mode_pvrow_index=fast_mode_pvrow_index,
fast_mode_segment_index=fast_mode_segment_index)
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Run fast mode
def fn_report(pvarray):
return (pvarray.ts_pvrows[1].back.get_param_weighted('qinc'))
qinc_fast = eng.run_fast_mode(fn_build_report=fn_report)
# Run full mode
report = eng.run_full_mode(fn_build_report=fn_report_example)
np.testing.assert_allclose(qinc_fast, 119.095285)
np.testing.assert_allclose(report['qinc_back'], 116.49050349491)
def pv_engine_run(data, pvarray_parameters, parallel=0):
if parallel > 0:
report = run_parallel_engine(MyReportBuilder,
pvarray_parameters,
data.index,
data.dni,
data.dhi,
data.zenith,
data.azimuth,
data.surface_tilt,
data.surface_azimuth,
data.albedo,
n_processes=6)
else:
rb = MyReportBuilder()
engine = PVEngine(pvarray_parameters)
engine.fit(data.index, data.dni, data.dhi, data.zenith, data.azimuth,
data.surface_tilt, data.surface_azimuth, data.albedo)
report = engine.run_all_timesteps(rb.build)
df_report = pd.DataFrame(report, index=data.index).fillna(0)
return df_report
def test_create_engine_with_rho_init(params, pvmodule_canadian):
"""Check that can create PV engine with rho initialization
from faoi functions"""
# Create inputs
pvarray = OrderedPVArray.init_from_dict(params)
irradiance = HybridPerezOrdered(rho_front=None, rho_back=None)
faoi_fn = faoi_fn_from_pvlib_sandia(pvmodule_canadian)
vfcalculator = VFCalculator(faoi_fn_front=faoi_fn, faoi_fn_back=faoi_fn)
# Create engine
engine = PVEngine.with_rho_initialization(pvarray, vfcalculator,
irradiance)
# Check that rho values are the ones calculated
np.testing.assert_allclose(engine.irradiance.rho_front, 0.02900688)
np.testing.assert_allclose(engine.irradiance.rho_back, 0.02900688)
def test_engine_variable_albedo(params, df_inputs_clearsky_8760):
"""Run PV engine calcs with variable albedo"""
n_points = 100 # limiting because circleci is taking very long
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray, irradiance_model=irradiance_model)
# Manage timeseries inputs
df_inputs = df_inputs_clearsky_8760[:n_points]
# Get MET data
timestamps = df_inputs.index
dni = df_inputs.dni.values
dhi = df_inputs.dhi.values
solar_zenith = df_inputs.solar_zenith.values
solar_azimuth = df_inputs.solar_azimuth.values
surface_tilt = df_inputs.surface_tilt.values
surface_azimuth = df_inputs.surface_azimuth.values
albedo = np.linspace(0.01, 1, num=n_points)
# Fit engine
eng.fit(timestamps, dni, dhi, solar_zenith, solar_azimuth, surface_tilt,
surface_azimuth, albedo)
# Run timestep
pvarray = eng.run_full_mode(fn_build_report=lambda pvarray: pvarray)
# Check the bifacial gain values
pvrow = pvarray.ts_pvrows[1]
bfg = (np.nansum(pvrow.back.get_param_weighted('qinc')) /
np.nansum(pvrow.front.get_param_weighted('qinc'))) * 100.
bfg_after_aoi = (np.nansum(pvrow.back.get_param_weighted('qabs')) /
np.nansum(pvrow.front.get_param_weighted('qabs'))) * 100.
expected_bfg = 14.973198
expected_bfg_after_aoi = 14.670709
np.testing.assert_allclose(bfg, expected_bfg)
np.testing.assert_allclose(bfg_after_aoi, expected_bfg_after_aoi)
def test_run_fast_mode_segments(params):
"""Test that PV engine works for timeseries fast mode and float inputs.
Value is very close to loop-like fast mode"""
# Discretize middle PV row's back side
params.update({'cut': {1: {'back': 5}}})
# Prepare some engine inputs
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(
params, param_names=irradiance_model.params)
fast_mode_pvrow_index = 1
fast_mode_segment_index = 2
# Create engine object
eng = PVEngine(pvarray,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index,
fast_mode_segment_index=fast_mode_segment_index)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_illum_pvrow'],
DNI)
# Define report function to grab irradiance from PV row segment
def fn_report(pvarray):
return (pvarray.ts_pvrows[1].back.list_segments[2].get_param_weighted(
'qinc'))
# Expected value for middle segment
qinc_expected = 116.572594
# Run fast mode for specific segment
qinc_segment = eng.run_fast_mode(fn_build_report=fn_report)
# Check results
np.testing.assert_allclose(qinc_segment, qinc_expected)
# Without providing segment index: the value should be the same as above
eng.fast_mode_segment_index = None
qinc_segment = eng.run_fast_mode(fn_build_report=fn_report)
# Check results
np.testing.assert_allclose(qinc_segment, qinc_expected)
def test_run_fast_mode_back_shading(params):
"""Test that PV engine works for timeseries fast mode and float inputs,
and when there's large direct shading on the back surface.
Value is very close to loop-style fast mode"""
params.update({
'gcr': 0.6,
'surface_azimuth': 270,
'surface_tilt': 120,
'solar_zenith': 70.
})
# Prepare some engine inputs
irradiance_model = HybridPerezOrdered()
pvarray = OrderedPVArray.init_from_dict(
params, param_names=irradiance_model.params)
fast_mode_pvrow_index = 1
fast_mode_segment_index = 0
# Create engine object
eng = PVEngine(pvarray,
irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index,
fast_mode_segment_index=fast_mode_segment_index)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Expected values
expected_qinc = 683.537153
# Fit engine
eng.fit(timestamps, DNI, DHI, params['solar_zenith'],
params['solar_azimuth'], params['surface_tilt'],
params['surface_azimuth'], params['rho_ground'])
def fn_report(pvarray):
return (pvarray.ts_pvrows[1].back.get_param_weighted('qinc'))
# By providing segment index
qinc = eng.run_fast_mode(fn_build_report=fn_report)
# Check results
np.testing.assert_allclose(qinc, expected_qinc)
# Without providing segment index
eng.fast_mode_segment_index = None
qinc = eng.run_fast_mode(fn_build_report=fn_report)
# Check results
np.testing.assert_allclose(qinc, expected_qinc)
def test_run_fast_mode_isotropic(params):
"""Test that PV engine works for timeseries fast mode and float inputs,
and using the isotropic irradiance model"""
# Prepare some engine inputs
irradiance_model = IsotropicOrdered()
pvarray = OrderedPVArray.init_from_dict(
params, param_names=irradiance_model.params)
fast_mode_pvrow_index = 1
fast_mode_segment_index = 0
# Create engine object
eng = PVEngine(pvarray, irradiance_model=irradiance_model,
fast_mode_pvrow_index=fast_mode_pvrow_index,
fast_mode_segment_index=fast_mode_segment_index)
# Irradiance inputs
timestamps = dt.datetime(2019, 6, 11, 11)
DNI = 1000.
DHI = 100.
# Fit engine
eng.fit(timestamps, DNI, DHI,
params['solar_zenith'], params['solar_azimuth'],
params['surface_tilt'], params['surface_azimuth'],
params['rho_ground'])
# Checks
np.testing.assert_almost_equal(eng.irradiance.direct['front_illum_pvrow'],
DNI)
# Expected value
qinc_expected = 122.73453
# Run fast mode
qinc = eng.run_fast_mode(
fn_build_report=lambda pvarray: (pvarray.ts_pvrows[1]
.back.get_param_weighted('qinc')))
# Check results
np.testing.assert_allclose(qinc, qinc_expected)
# Without providing segment index
eng.fast_mode_segment_index = None
qinc = eng.run_fast_mode(
fn_build_report=lambda pvarray: (pvarray.ts_pvrows[1]
.back.get_param_weighted('qinc')))
# Check results
np.testing.assert_allclose(qinc, qinc_expected)
def test_check_tilt_zero_discontinuity():
"""
Before version 1.5.2, surface_tilt=0 with certain combinations of
surface_azimuth and axis_azimuth showed anomolous behavior where
the irradiance at zero tilt was significantly different from the
irradiance at very small but nonzero tilts. Additionally, the
calculated VF matrix could have values outside [0, 1]. See GH #125
"""
# expected value calculated for surface_tilt=0.001, so should
# not be significantly different from result for surface_tilt=0
rear_qinc_expected = 76.10
timestamps = np.array([dt.datetime(2019, 6, 1, 10)])
solar_azimuth = np.array([135])
solar_zenith = np.array([45])
dni = np.array([200])
dhi = np.array([400])
albedo = np.array([0.2])
surface_tilt = np.array([0.0])
# the discontinuity did not occur for all combinations of
# (surface_azimuth, axis_azimuth), so test all four "primary" pairs:
for surface_azimuth in [90, 270]:
surface_azimuth = np.array([surface_azimuth])
for axis_azimuth in [0, 180]:
params = dict(n_pvrows=3,
axis_azimuth=axis_azimuth,
pvrow_height=2,
pvrow_width=1,
gcr=0.4)
pvarray = OrderedPVArray.init_from_dict(params)
eng = PVEngine(pvarray)
eng.fit(timestamps, dni, dhi, solar_zenith, solar_azimuth,
surface_tilt, surface_azimuth, albedo)
# Run simulation and get output
eng.run_full_mode()
out = pvarray.ts_pvrows[1].back.get_param_weighted('qinc')
assert np.all(pvarray.ts_vf_matrix >= 0)
assert np.all(pvarray.ts_vf_matrix <= 1)
assert rear_qinc_expected == pytest.approx(out[0], abs=1e-2)
def pvfactors_engine_run(data, pvarray_parameters, parallel=0, mode='full'):
"""My wrapper function to launch the pvfactors engine in parallel. It is mostly for Windows use.
In Linux you can directly call run_parallel_engine. It uses MyReportBuilder to generate the output.
Args:
data (pandas DataFrame): The data to fit the model.
pvarray_parameters (dict): The pvfactors dict describing the simulation.
parallel (int, optional): Number of threads to launch. Defaults to 0 (just calls PVEngine.run_all_timesteps)
mode (str): full or fast depending on the type of back irraadiances. See pvfactors doc.
Returns:
pandas DataFrame: The results of the simulation, as desired in MyReportBuilder.
"""
n, row = _get_cut(pvarray_parameters['cut'])
rb = Report(n, row)
if parallel > 1:
report = run_parallel_engine(rb,
pvarray_parameters,
data.index,
data.dni,
data.dhi,
data.zenith,
data.azimuth,
data.surface_tilt,
data.surface_azimuth,
data.albedo,
n_processes=parallel)
else:
pvarray = OrderedPVArray.init_from_dict(pvarray_parameters)
engine = PVEngine(pvarray)
engine.fit(data.index, data.dni, data.dhi, data.zenith, data.azimuth,
data.surface_tilt, data.surface_azimuth, data.albedo,
data.ghi)
if mode == 'full': report = engine.run_full_mode(rb.build)
else:
report = engine.run_fast_mode(rb.build,
pvrow_index=0,
segment_index=0)
df_report = pd.DataFrame(report, index=data.index).fillna(0)
return df_report