def test_histogram_data_nested():
# Testing vals
dir_vals = [0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 288]
spd_vals = dir_vals
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
# Init simple example w segs == bin num
w = WindRose(dir_data, spd_data, 4)
#w.legend_parameters = LegendParameters(segment_count=5)
w.frequency_hours = 1
# Bin values to divide into colors
# 315-45: [10, 10, 10]; 2 intervals, [10, 10, 10]
# 45-135: [85, 90, 90, 90, 95]; 3 intervals, [85, 90, 90, 90, 95]
# 135-225: [170]; 1 intervals, [170];
# 225-315: [285, 288]; 2 intervals, [285, 288]
# interval_num: [2, 3, 1, 2]
chk_histstack = [
[10, 10, 10],
[85, 90, 90, 90, 95],
[170.],
[285, 288]]
# Testing
histstack = WindRose._histogram_data_nested(w.histogram_data, 1)
for chkh, h in zip(chk_histstack, histstack):
for c, _h in zip(chkh, h):
assert abs(c - _h) <= 1e-10
# Init complex dir set divided by 4
w = WindRose(dir_data, spd_data, 4)
w.frequency_hours = 2
# Bin values to divide into colors
# 315-45: [10, 10, 10]; 2 intervals, [10, 10]
# 45-135: [85, 90, 90, 90, 95]; 3 intervals, [87.5, 90, 95. ]
# 135-225: [170]; 1 intervals, [170]
# 225-315: [285, 288]; 2 intervals, [286.5]
# interval_num: [2, 3, 1, 2]
chk_histstack = [
[10, 10],
[87.5, 90, 95.],
[170.],
[286.5]]
# Testing
histstack = WindRose._histogram_data_nested(w.histogram_data, 2)
for chkh, h in zip(chk_histstack, histstack):
for c, _h in zip(chkh, h):
assert abs(c - _h) <= 1e-10
def test_frequency_intervals():
"""Test the distance of frequency_intervals"""
# Plot windrose
epw_path = os.path.join(os.getcwd(), 'tests/fixtures/epw/tokyo.epw')
epw = EPW(epw_path)
w = WindRose(epw.wind_direction, epw.wind_speed, 3)
w.show_zeros = False
w.show_freq = False
w.frequency_hours = 200.0
test_freq_int = int(math.ceil(w.real_freq_max / 200.0))
assert w.frequency_intervals_compass == pytest.approx(test_freq_int, abs=1e-10)
assert w.frequency_intervals_mesh == pytest.approx(test_freq_int, abs=1e-10)
# Test changing interals from 18
# Reduce
w.frequency_intervals_compass = 10.0
test_freq_int = 10.0
assert w.frequency_maximum == pytest.approx(10 * 200.0, abs=1e-10)
assert w.frequency_intervals_compass == pytest.approx(test_freq_int, abs=1e-10)
assert w.frequency_intervals_mesh == pytest.approx(test_freq_int, abs=1e-10)
# Check that resetting frequency_max works
w._frequency_intervals_compass = None
# w.real_freq_max: 4406
chk_max = int(math.ceil(4406 / 200)) * 200
assert w.frequency_maximum == pytest.approx(chk_max, abs=1e-10)
assert w.frequency_intervals_compass == pytest.approx(23, abs=1e-10)
assert w.frequency_intervals_mesh == pytest.approx(23, abs=1e-10)
def test_wind_polygons():
"""Test colors for different windrose types."""
# Testing vals
dir_vals = [0, 0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 310]
spd_vals = [0, 0, 0, 0, 1, 145, 189, 15, 10, 150, 300, 259, 100, 5, 301]
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
# Bin values to divide into 6 intervals = 10 - 310 = 300 / 6 = 50 m/s
# intervals: [1-50, 51-100, 101-150, 151-200, 201-250, 251-300, 301-350]
#
# [[1], [], [145], [189], [], []], # 0-49, 100-149, 150-199
# [[10, 15], [], [150], [], [], [259, 300]], # 0-49, 150-199,, 250-299
# [[], [100], [], [], [], []], # 100-149
# [[5], [], [], [], [], [301]] # 0-49, 300-349
# Check freq=True, zeros=False
w = WindRose(dir_data, spd_data, 4)
w.show_freq, w.show_zeros = True, False
w.frequency_hours = 2
w.legend_parameters = LegendParameters(segment_count=6)
chk_poly_num = sum([3, 3, 1, 2])
assert chk_poly_num == len(w.windrose_lines)
# For averaged
w.show_freq = False
chk_poly_num = 4
w.colored_mesh
assert chk_poly_num == len(w.windrose_lines)
all_orient_line = []
all_freq_line = []
all_legends = []
all_title = []
all_calm_hours = []
all_histograms = []
# Calculate _max_freq_lines_ if it's not already set, to use to
# determine spacing for multiple plots.
if len(_data) > 1 and _max_freq_lines_ is None:
max_freqs = []
for i, _data_item in enumerate(_data):
win_dir = _wind_direction
w = WindRose(win_dir, _data_item, _dir_count_)
if _freq_hours_ is not None:
w.frequency_hours = _freq_hours_
if _freq_dist_ is not None:
w.frequency_spacing_distance = _freq_dist_
max_freqs.append(w.frequency_intervals_compass)
_max_freq_lines_ = max(max_freqs)
# Plot the windroses
for i, speed_data in enumerate(_data):
# Make the windrose
win_dir = _wind_direction
windrose = WindRose(win_dir, speed_data, _dir_count_)
if len(legend_par_) > 0:
try: # sense when several legend parameters are connected
lpar = legend_par_[i]
def test_color_array():
"""Test colors for different windrose types."""
# Testing vals
dir_vals = [0, 0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 310]
spd_vals = [0, 0, 0, 0, 1, 145, 189, 15, 10, 150, 300, 259, 100, 5, 301]
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
data_step = 50.0
min_val = 1
# Bin values to divide into 6 intervals = 10 - 310 = 300 / 6 = 50 m/s
# intervals: [1-50, 51-100, 101-150, 151-200, 201-250, 251-300, 301-350]
#
# [[1], [], [145], [189], [], [], # 0-49, 100-149, 150-199
# [[10, 15], [], [150], [], [], [259, 300]], # 0-49, 150-199,, 250-299
# [[], [100], [], [], [], []], # 100-149
# [[5], [], [], [], [], [301]] # 0-49, 300-349
# Check freq=True, zeros=False
w = WindRose(dir_data, spd_data, 4)
w.show_freq, w.show_zeros = True, False
w.frequency_hours = 2
w.legend_parameters = LegendParameters(segment_count=7)
w.colored_mesh
# color index corresponds to hourly interval indices
# [1-50, 51-100, 101-150, 151-200, 201-250, 251-300, 301-350]
chk_color_array = [0, 2, 3, 0, 2, 5, 1, 0, 5]
chk_color_array = [(c * data_step) + min_val for c in chk_color_array]
assert len(chk_color_array) == len(w._color_array)
for cc, c in zip(chk_color_array, w._color_array):
assert abs(cc - c) < 1e-10
# Check freq=True, zeros=True
# Modify range for easier calcs
dir_vals = [0, 0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 310]
spd_vals = [0, 0, 0, 0, 1, 145, 189, 15, 10, 149, 299, 259, 99, 5, 300]
zero_spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
zero_dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
zero_data_step = 50
zero_min_val = 0
w = WindRose(zero_dir_data, zero_spd_data, 4)
w.show_freq, w.show_zeros = True, True
w.frequency_hours = 2
w.legend_parameters = LegendParameters(segment_count=7)
w.colored_mesh
# color index corresponds to hourly interval indices
chk_color_array = [1, 3, 4, 1, 3, 6, 2, 1, 6]
chk_color_array = [(c * zero_data_step) + zero_min_val for c in chk_color_array]
chk_color_array += [0, 0, 0, 0]
assert len(chk_color_array) == len(w._color_array)
for cc, c in zip(chk_color_array, w._color_array):
assert abs(cc - c) < 1e-10
# Check freq=False, zeros=False
w = WindRose(dir_data, spd_data, 4)
w.show_freq, w.show_zeros = False, False
w.frequency_hours = 2
w.legend_parameters = LegendParameters(segment_count=7)
w.colored_mesh
# color index corresponds to hourly interval indices
chk_color_array = [111 + 2/3, 146.8, 100, 153]
assert len(chk_color_array) == len(w._color_array)
for cc, c in zip(chk_color_array, w._color_array):
assert abs(cc - c) < 1e-10
# Check freq=False, zeros=True
w = WindRose(dir_data, spd_data, 4)
w.show_freq, w.show_zeros = False, True
w.frequency_hours = 2
w.legend_parameters = LegendParameters(segment_count=7)
w.colored_mesh
# color index corresponds to hourly interval indices
chk_color_array = [111 + 2/3, 146.8, 100, 153]
chk_color_array += [0, 0, 0, 0]
assert len(chk_color_array) == len(w._color_array)
for cc, c in zip(chk_color_array, w._color_array):
assert abs(cc - c) < 1e-10
def test_histogram_data_nested():
# Testing vals
dir_vals = [0, 0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 310]
spd_vals = [0, 0, 0, 0, 1, 145, 189, 15, 10, 150, 299, 259, 100, 5, 301]
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
# Init complex dir set divided by 4, and 2 hourly intervals
w = WindRose(dir_data, spd_data, 4)
w.frequency_hours = 2
w.legend_parameters.segment_count = 7
# Bin values to divide into 6 intervals = 10 - 310 = 300 / 6 = 50 m/s
# intervals: [1-50, 51-100, 101-150, 151-200, 201-250, 251-300, 301-350]
#
# each frequency interval = 1/6 ~ 0.166 of total speed
# 315-45: [1, 145, 189]; 3 bands [0-49, 100-149, 150-199]
# 45-135: [15, 10, 150, 299, 259]; 3 bands, [0-49, 150-199,, 250-299]
# 135-225: [100]; 1 bands, [100-149]
# 225-315: [5, 300]; 2 bands, [0-49, 300, 249]
# interval_num: [3, 3, 1, 2]
chk_histstack = [
[[1], [], [145], [189], [], []], # 0-49, 100-149, 150-199
[[10, 15], [], [150], [], [], [259, 299]], # 0-49, 150-199,, 250-299
[[], [100], [], [], [], []], # 100-149
[[5], [], [], [], [], [301]]] # 0-49, 300-349
# Testing
speeds = [val for bin in w.histogram_data for val in bin]
min_speed, max_speed = min(speeds), max(speeds)
histstack, bins = WindRose._histogram_data_nested(
w.histogram_data, (min_speed, max_speed), w.legend_parameters.segment_count)
# Check
assert len(chk_histstack) == len(histstack)
for cbins, bins in zip(chk_histstack, histstack):
assert len(cbins) == len(bins)
for cbin, bin in zip(cbins, bins):
assert len(cbin) == len(bin)
for cval, val in zip(cbin, bin):
assert abs(cval - val) <= 1e-10, (cval, val)
# Check with zeros
w.show_zeros = True
w.frequency_hours = 1.0
w.frequency_spacing_distance = 25.0
hypot_dist = w.frequency_spacing_hypot_distance
assert abs(w._zero_mesh_radius - hypot_dist) < 1e-10
# Check with zeros 2
# Testing vals
dir_vals = [0, 0, 0, 0, 0, 0, 10, 10, 10, 85, 90, 90, 90, 95, 170, 285, 310]
spd_vals = [0, 0, 0, 0, 0, 0, 1, 145, 189, 15, 10, 150, 299, 259, 100, 5, 301]
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
# Init complex dir set divided by 4, and 2 hourly intervals
w = WindRose(dir_data, spd_data, 4)
w.show_zeros = True
w.frequency_hours = 1.0
w.frequency_spacing_distance = 25.0
hypot_dist = w.frequency_spacing_hypot_distance * 1.5
assert abs(w._zero_mesh_radius - hypot_dist) < 1e-10
def test_windrose_frequency_distribution():
"""Test frequency distribution"""
# Testing mesh scaling
dir_vals = [1, 2, 3, 4, 5, 6,
90, 91, 92, 93, 94, 95,
180, 181, 182, 183, 184, 185,
270, 271, 272, 273, 274, 275]
spd_vals = dir_vals
# Make into fake data collections
a_per = AnalysisPeriod(6, 21, 12, 6, 21, 13)
dates = [DateTime(6, 21, i) for i in range(len(dir_vals))]
spd_header = Header(Speed(), 'm/s', a_per)
dir_header = Header(GenericType('Direction', 'deg'), 'deg', a_per)
spd_data = HourlyDiscontinuousCollection(spd_header, spd_vals, dates)
dir_data = HourlyDiscontinuousCollection(dir_header, dir_vals, dates)
# Init simple dir set divided by 4
w = WindRose(dir_data, spd_data, 4)
w.frequency_hours = 3
assert w.frequency_intervals_mesh == 2
assert w.real_freq_max == 6
assert w.frequency_maximum == 6
freqs = WindRose._histogram_data_nested(w.histogram_data, 3)
north_hbin = freqs[0]
assert north_hbin[0] == 2.0 # [1, 2, 3]
assert north_hbin[1] == 5.0 # [4, 5, 6]
assert north_hbin[0] == sum(w.histogram_data[0][:3]) / 3.0
# Test w/ epw
epw_path = os.path.join(os.getcwd(), 'tests/fixtures/epw/tokyo.epw')
epw = EPW(epw_path)
# Test 16 directions
w = WindRose(epw.wind_direction, epw.wind_speed, 3)
w.show_zeros = False
w.show_freq = True
# w.real_freq_max: 4406
# w.frequency_maximum: 4600
# Test w/ no stacking
w.frequency_hours = 4600 # 1 bin
ytick_num = w.frequency_intervals_mesh
assert ytick_num == 1
freqs = WindRose._histogram_data_nested(w.histogram_data, 4600)
hbin = freqs[0]
test_val = sum(w.histogram_data[0]) / len(w.histogram_data[0])
assert hbin[0] == pytest.approx(test_val, abs=1e-10)
# Test w/ stacking
w.frequency_hours = 5 # 1 bin
h = w.frequency_hours
freqs = WindRose._histogram_data_nested(w.histogram_data, h)
hbin = freqs[0]
sort_hist_bar = sorted(w.histogram_data[0])
test_val = sum(sort_hist_bar[:5]) / 5.0
assert hbin[0] == pytest.approx(test_val, abs=1e-10)
