import pytest import scipy.stats import numpy as np from questplus.qp import QuestPlus, QuestPlusWeibull, QuestPlusThurstone from questplus import _constants def test_threshold(): """ Watson 2017, Example 1: "Estimation of contrast threshold {1, 1, 2}" """ threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() expected_contrasts = [-18, -22, -25, -28, -30, -22, -13, -15, -16, -18, -19, -20, -21, -22, -23, -19, -20, -20, -18, -18, -19, -17, -17, -18, -18, -18, -19, -19, -19, -19, -19, -19] responses = ['Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct'] expected_mode_threshold = -20 stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) for expected_contrast, response in zip(expected_contrasts, responses): assert q.next_stim == dict(intensity=expected_contrast) q.update(stim=q.next_stim, outcome=dict(response=response)) assert np.allclose(q.param_estimate['threshold'], expected_mode_threshold) def test_threshold_slope(): """ Watson 2017, "Implementation" """ threshold = np.arange(-40, 0 + 1) slope = np.arange(2, 5 + 1) guess = 0.5 lapse = 0.02 contrasts = threshold.copy() expected_mode_threshold = -20 expected_mode_slope = 3 expected_contrasts = [-18, -22, -12, -13, -15, -16, -17, -18, -19, -21, -24, -25, -26, -27, -21, -22, -23, -20, -20, -20, -21, -19, -19, -20, -20, -20, -19, -19, -19, -19, -19, -20, -20, -20, -20, -20, -21, -21, -20, -20, -20, -19, -18, -18, -19, -19, -19, -18, -18, -17, -18, -18, -18, -17, -17, -17, -17, -17, -17, -17, -18, -18, -18, -18] responses = ['Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct'] stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) for expected_contrast, response in zip(expected_contrasts, responses): assert q.next_stim == dict(intensity=expected_contrast) q.update(stim=q.next_stim, outcome=dict(response=response)) assert np.allclose(q.param_estimate['threshold'], expected_mode_threshold) assert np.allclose(q.param_estimate['slope'], expected_mode_slope) def test_threshold_slope_lapse(): """ Watson 2017, Example 2: "Estimation of contrast threshold, slope, and lapse {1, 3, 2}" """ expected_mode_threshold = -20 expected_mode_slope = 5 expected_mode_lapse = 0.04 contrasts = np.arange(-40, 0 + 1) slope = np.arange(2, 5 + 1) lower_asymptote = 0.5 lapse_rate = np.arange(0, 0.04 + 0.01, 0.01) stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=contrasts, slope=slope, lower_asymptote=lower_asymptote, lapse_rate=lapse_rate) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' expected_contrasts = [-18, -22, -25, -28, -30, -21, -12, -14, -15, -16, -18, -19, -20, -22, -23, -18, -19, -16, -17, -17, -18, -18, -18, -19, -19, -17, -17, -18, -18, -18, -18, -19, -19, -20, -18, -18, -18, -18, -18, -19, -19, -19, -19, -20, -18, -18, -18, -18, -18, -18, -18, -18, -18, -19, -19, -19, -20, -18, -18, -18, -18, -19, -19, -19, -18, -18, -18, -18, -19, -19, -19, -19, -19, -19, -19, -19, -20, -21, -19, -19, -19, -19, -19, -19, -19, -18, -19, -19, -16, -16, -16, -16, -16, -16, 0, 0, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -21, -21, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19, -19] responses = ['Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct'] q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) for expected_contrast, response in zip(expected_contrasts, responses): assert q.next_stim == dict(intensity=expected_contrast) q.update(stim=q.next_stim, outcome=dict(response=response)) assert np.allclose(q.param_estimate['threshold'], expected_mode_threshold) assert np.allclose(q.param_estimate['slope'], expected_mode_slope) assert np.allclose(q.param_estimate['lapse_rate'], expected_mode_lapse) def test_mean_sd_lapse(): """ Watson 2017, Example 3: "Estimation of mean, standard deviation, and lapse {1, 3, 2}" """ expected_mode_mean = 0 expected_mode_sd = 4 expected_mode_lapse_rate = 0.01 # Stimulus domain. orientation = np.arange(-10, 10+1) stim_domain = dict(intensity=orientation) # Parameter domain. orientation = np.arange(-5, 5+1) sd = np.arange(1, 10+1) lapse_rate = np.arange(0, 0.04 + 0.01, 0.01) param_domain = dict(mean=orientation, sd=sd, lapse_rate=lapse_rate) # Outcome domain. outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'norm_cdf_2' scale = 'linear' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' expected_orientations = [ 0, -1, 2, -2, 2, 3, 3, -5, 3, 3, 3, 3, -7, -7, -7, -7, 4, -7, 4, 4, 4, -8, -8, -8, 4, -8, 4, -8, -7, 4, -7, -7, 4, -7, 4, -7, -6, 3, -4, -5, 3, 3, -7, 2, 2, 1, 0, -7, -6, -6, -6, 2, -5, 2, -6, -5, 3, -5, -5, 3, 4, -5, -5, 4, -4, -4, -4, -4, 4, 3, 3, 4, 5, 5, -5, 5, -5, -5, 4, 5, 5, -5, -4, -4, -4, -4, 5, -4, -4, -3, 4, 4, 5, 5, -4, 5, -4, -4, 5, 4, 5, 5, -5, 5, -4, -4, -4, -4, 5, -4, 5, 4, -4, 4, -4, -4, -4, 4, -4, -4, -4, 4, 4, -4, 4, 4, 4, -4] responses = ['Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Incorrect', 'Incorrect', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect'] q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) for expected_orientation, response in zip(expected_orientations, responses): next_stim = q.next_stim assert next_stim == dict(intensity=expected_orientation) q.update(stim=next_stim, outcome=dict(response=response)) assert np.allclose(q.param_estimate['mean'], expected_mode_mean) assert np.allclose(q.param_estimate['sd'], expected_mode_sd) assert np.allclose(q.param_estimate['lapse_rate'], expected_mode_lapse_rate) def test_spatial_contrast_sensitivity(): """ Watson 2017, Example 4: "Contrast sensitivity function {2, 3, 2}" """ spatial_freqs = np.arange(0, 40 + 2, 2) contrasts = np.arange(-50, 0 + 2, 2) temporal_freq = 0 # Stimulus domain. stim_domain = dict(contrast=contrasts, spatial_freq=spatial_freqs, temporal_freq=temporal_freq) # Parameter domain. min_threshs = np.arange(-50, -30 + 2, 2) c0s = np.arange(-60, -40 + 2, 2) cfs = np.arange(0.8, 1.6 + 0.2, 0.2) cw = 0 slope = 3 lower_asymptote = 0.5 lapse_rate = 0.01 param_domain = dict(min_thresh=min_threshs, c0=c0s, cf=cfs, cw=cw, slope=slope, lower_asymptote=lower_asymptote, lapse_rate=lapse_rate) # Outcome domain. outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'csf' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' expected_mode_min_thresh = -32 expected_mode_c0 = -56 expected_mode_cf = 1.4 expected_contrasts = [0, -4, 0, 0, -38, 0, -40, 0, -26, -26, 0, -36, -36, 0, -26, -26, -2, -26, -6, -26, 0, -26, 0, -26, -32, -32, -34, -34, 0, -26, 0, -26] expected_spatial_freqs = [40, 40, 34, 36, 0, 38, 0, 38, 18, 18, 40, 0, 0, 40, 20, 20, 40, 22, 40, 20, 40, 18, 40, 18, 0, 0, 0, 0, 40, 18, 38, 18] responses = ['Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct'] q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) for x in zip(expected_contrasts, expected_spatial_freqs, responses): expected_contrast, expected_spatial_freq, response = x assert q.next_stim == dict(contrast=expected_contrast, spatial_freq=expected_spatial_freq, temporal_freq=0) q.update(stim=q.next_stim, outcome=dict(response=response)) assert np.allclose(q.param_estimate['min_thresh'], expected_mode_min_thresh) assert np.allclose(q.param_estimate['c0'], expected_mode_c0) assert np.allclose(q.param_estimate['cf'], expected_mode_cf) def test_thurstone_scaling(): """ Watson 2017, Example 6: "Thurstone scaling {2, 3, 2}" """ stim_magnitudes = np.arange(0, 1+0.1, 0.1) perceptual_scale_maxs = np.arange(1, 10+1) thresholds = np.arange(0, 0.9+0.1, 0.1) powers = np.arange(0.1, 1+0.1, 0.1) # Due to differences in rounding, the order of stimuli (1 or 2) is swapped on some trials # compared to the paper. We therefore have to swap the example response as well. # # We're only testing the first 22 trials here. responses = ['Second'] * 6 responses.extend(['Second']) # rounding difference responses.extend(['Second'] * 13) responses.extend(['Second']) # rounding difference responses.extend(['First']) expected_stims = [ {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.7}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.6}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.5}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.4}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.0}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.4}, {'physical_magnitude_stim_1': 0.0, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.5, 'physical_magnitude_stim_2': 1.0}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.5, 'physical_magnitude_stim_2': 1.0}, {'physical_magnitude_stim_1': 0.5, 'physical_magnitude_stim_2': 1.0}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.5, 'physical_magnitude_stim_2': 1.0}, {'physical_magnitude_stim_1': 0.6, 'physical_magnitude_stim_2': 1.0}, {'physical_magnitude_stim_1': 0.2, 'physical_magnitude_stim_2': 0.3}, {'physical_magnitude_stim_1': 0.6, 'physical_magnitude_stim_2': 1.0}, ] qp = QuestPlusThurstone( physical_magnitudes_stim_1=stim_magnitudes, physical_magnitudes_stim_2=stim_magnitudes, thresholds=thresholds, powers=powers, perceptual_scale_maxs=perceptual_scale_maxs ) for trial_idx, x in enumerate(zip(expected_stims, responses)): expected_stim, response = x expected_stim_1 = expected_stim['physical_magnitude_stim_1'] expected_stim_2 = expected_stim['physical_magnitude_stim_2'] next_stim_1 = qp.next_stim['physical_magnitude_stim_1'] next_stim_2 = qp.next_stim['physical_magnitude_stim_2'] if trial_idx in (6, 20, 21): # Rounding errors make the algorithm behave differently on different platforms. if ( np.isclose(expected_stim_1, next_stim_2) and np.isclose(expected_stim_2, next_stim_1) ): expected_stim_1, expected_stim_2 = expected_stim_2, expected_stim_1 response = 'First' if response == 'Second' else 'Second' assert np.isclose(next_stim_1, expected_stim_1) assert np.isclose(next_stim_2, expected_stim_2) qp.update(stim=qp.next_stim, response=response) def test_weibull(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() expected_contrasts = [-18, -22, -25, -28, -30, -22, -13, -15, -16, -18, -19, -20, -21, -22, -23, -19, -20, -20, -18, -18, -19, -17, -17, -18, -18, -18, -19, -19, -19, -19, -19, -19] responses = ['Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Incorrect', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct', 'Correct'] expected_mode_threshold = -20 stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) q_weibull = QuestPlusWeibull(intensities=stim_domain['intensity'], thresholds=param_domain['threshold'], slopes=param_domain['slope'], lower_asymptotes=param_domain['lower_asymptote'], lapse_rates=param_domain['lapse_rate'], responses=outcome_domain['response'], stim_scale=scale) for expected_contrast, response in zip(expected_contrasts, responses): assert q.next_stim['intensity'] == q_weibull.next_intensity assert q_weibull.next_intensity == expected_contrast q.update(stim=q.next_stim, outcome=dict(response=response)) q_weibull.update(intensity=q_weibull.next_intensity, response=response) assert np.allclose(q.param_estimate['threshold'], expected_mode_threshold) def test_eq(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' q1 = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) q2 = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) # Add some random responses. q1.update(stim=q1.next_stim, outcome=dict(response='Correct')) q1.update(stim=q1.next_stim, outcome=dict(response='Incorrect')) q2.update(stim=q2.next_stim, outcome=dict(response='Correct')) q2.update(stim=q2.next_stim, outcome=dict(response='Incorrect')) assert q1 == q2 def test_json(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) # Add some random responses. q.update(stim=q.next_stim, outcome=dict(response='Correct')) q.update(stim=q.next_stim, outcome=dict(response='Incorrect')) q_dumped = q.to_json() q_loaded = QuestPlus.from_json(q_dumped) assert q_loaded == q q_loaded.update(stim=q_loaded.next_stim, outcome=dict(response='Correct')) def test_json_rng(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_n_entropy' param_estimation_method = 'mode' random_seed = 5 stim_selection_options = dict(n=3, random_seed=random_seed) q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method, stim_selection_options=stim_selection_options) q2 = QuestPlus.from_json(q.to_json()) rand = q._rng.random_sample(10) rand2 = q2._rng.random_sample(10) assert np.allclose(rand, rand2) def test_marginal_posterior(): contrasts = np.arange(-40, 0 + 1) slope = np.arange(2, 5 + 1) lower_asymptote = (0.5,) lapse_rate = np.arange(0, 0.04 + 0.01, 0.01) stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=contrasts, slope=slope, lower_asymptote=lower_asymptote, lapse_rate=lapse_rate) outcome_domain = dict(response=['Correct', 'Incorrect']) func = 'weibull' stim_scale = 'dB' q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=func, stim_scale=stim_scale) marginal_posterior = q.marginal_posterior assert np.allclose(marginal_posterior['threshold'], np.ones(len(contrasts)) / len(contrasts)) assert np.allclose(marginal_posterior['slope'], np.ones(len(slope)) / len(slope)) assert np.allclose(marginal_posterior['lower_asymptote'], np.ones(len(lower_asymptote)) / len(lower_asymptote)) assert np.allclose(marginal_posterior['lapse_rate'], np.ones(len(lapse_rate)) / len(lapse_rate)) def test_prior_for_unknown_parameter(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' prior = dict(Foo=[1, 2, 3]) with pytest.raises(ValueError): q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method, prior=prior) def test_prior_for_parameter_subset(): threshold = np.arange(-40, 0 + 1) slopes = np.linspace(start=2, stop=4, num=11) guess, lapse = 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slopes, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_entropy' param_estimation_method = 'mode' threshold_prior = scipy.stats.norm.pdf(contrasts, loc=-20, scale=10) threshold_prior /= threshold_prior.sum() slope_prior = scipy.stats.norm.pdf(slopes, loc=3, scale=0.5) slope_prior /= slope_prior.sum() prior = dict(threshold=threshold_prior, slope=slope_prior) q = QuestPlus(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method, prior=prior) assert np.allclose(threshold_prior, q.prior.sum(dim=['slope', 'lower_asymptote', 'lapse_rate']).values) assert np.allclose(slope_prior, q.prior.sum(dim=['threshold', 'lower_asymptote', 'lapse_rate']).values) assert np.isclose(1, q.prior.sum(dim=['threshold', 'slope', 'lapse_rate']).sum()) assert np.isclose(1, q.prior.sum(dim=['threshold', 'slope', 'lower_asymptote']).sum()) def test_stim_selection_options(): threshold = np.arange(-40, 0 + 1) slope, guess, lapse = 3.5, 0.5, 0.02 contrasts = threshold.copy() stim_domain = dict(intensity=contrasts) param_domain = dict(threshold=threshold, slope=slope, lower_asymptote=guess, lapse_rate=lapse) outcome_domain = dict(response=['Correct', 'Incorrect']) f = 'weibull' scale = 'dB' stim_selection_method = 'min_n_entropy' param_estimation_method = 'mode' common_params = dict(stim_domain=stim_domain, param_domain=param_domain, outcome_domain=outcome_domain, func=f, stim_scale=scale, stim_selection_method=stim_selection_method, param_estimation_method=param_estimation_method) stim_selection_options = None q = QuestPlus(**common_params, stim_selection_options=stim_selection_options) expected = dict(n=_constants.DEFAULT_N, max_consecutive_reps=_constants.DEFAULT_MAX_CONSECUTIVE_REPS, random_seed=_constants.DEFAULT_RANDOM_SEED) assert expected == q.stim_selection_options stim_selection_options = dict(n=5) q = QuestPlus(**common_params, stim_selection_options=stim_selection_options) expected = dict(n=5, max_consecutive_reps=_constants.DEFAULT_MAX_CONSECUTIVE_REPS, random_seed=_constants.DEFAULT_RANDOM_SEED) assert expected == q.stim_selection_options stim_selection_options = dict(max_consecutive_reps=4) q = QuestPlus(**common_params, stim_selection_options=stim_selection_options) expected = dict(n=_constants.DEFAULT_N, max_consecutive_reps=4, random_seed=_constants.DEFAULT_RANDOM_SEED) assert expected == q.stim_selection_options stim_selection_options = dict(random_seed=999) q = QuestPlus(**common_params, stim_selection_options=stim_selection_options) expected = dict(n=_constants.DEFAULT_N, max_consecutive_reps=_constants.DEFAULT_MAX_CONSECUTIVE_REPS, random_seed=999) assert expected == q.stim_selection_options stim_selection_options = dict(n=5, max_consecutive_reps=4, random_seed=999) q = QuestPlus(**common_params, stim_selection_options=stim_selection_options) expected = stim_selection_options.copy() assert expected == q.stim_selection_options def test_weibull_prior(): intensities = np.linspace(-10, 0) thresholds = intensities.copy() slopes = [3.5] lower_asymptotes = [0.01] lapse_rates = [0.01] prior_val = scipy.stats.norm.pdf(intensities, loc=-5, scale=0.2) prior_val /= prior_val.sum() prior = dict(threshold=prior_val) q = QuestPlusWeibull(intensities=intensities, thresholds=thresholds, slopes=slopes, lower_asymptotes=lower_asymptotes, lapse_rates=lapse_rates, prior=prior) assert np.allclose(q.prior.squeeze().values, prior_val) if __name__ == '__main__': test_threshold() test_threshold_slope() test_threshold_slope_lapse() test_mean_sd_lapse() test_spatial_contrast_sensitivity() test_weibull() test_eq() test_json() test_json_rng() test_marginal_posterior() test_prior_for_unknown_parameter() test_prior_for_parameter_subset() test_stim_selection_options() test_weibull_prior()