-
Notifications
You must be signed in to change notification settings - Fork 27
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
1 parent
6357417
commit 85b934f
Showing
1 changed file
with
108 additions
and
108 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,108 +1,108 @@ | ||
| import os | ||
| import unittest | ||
|
|
||
| import numpy as np | ||
| # TODO: Currently silenced as transitioning away from Tensorflow. # import tensorflow as tf | ||
|
|
||
| # TODO: Currently silenced as transitioning away from Tensorflow. # from pyqsp import qsp_models | ||
|
|
||
| class Test_qsp_models(unittest.TestCase): | ||
| ''' | ||
| Unit tests for qsp_models component of pyqsp | ||
| ''' | ||
| def setUp(self): | ||
| ''' | ||
| Do imports here, so that this module can be entirely optional | ||
| ''' | ||
| if self.is_enabled(): | ||
| pass | ||
| else: | ||
| print( | ||
| "[pyqsp.test] Skipping qsp_model tests: export PYQSP_TEST_QSP_MODELS=1 to enable these tests") | ||
|
|
||
| def is_enabled(self): | ||
| enabled = 'PYQSP_TEST_QSP_MODELS' in os.environ | ||
| return enabled | ||
|
|
||
| def xtest_hsim1(self): | ||
| if not self.is_enabled(): | ||
| return | ||
| t = 3 | ||
| poly_deg = 6 | ||
| def f(x): return np.cos(t * x) | ||
| model = qsp_models.construct_qsp_model(poly_deg) | ||
|
|
||
| # The intput theta training values | ||
| th_in = np.arange(0, np.pi, np.pi / 50) | ||
| th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # The desired real part of p(x) which is the upper left value in the unitary of the qsp sequence | ||
| # We also want that Re[q(x)] = 0 | ||
| expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape)] | ||
|
|
||
| model = qsp_models.construct_qsp_model(poly_deg) | ||
| history = model.fit(x=th_in, y=expected_outputs, | ||
| epochs=1000, verbose=0) | ||
| # plot_loss(history) | ||
| # plot_qsp_response(f, model) | ||
| response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| model, return_all=True) | ||
| desired = f(np.cos(all_th)) | ||
| assert abs(response - desired).mean() < 0.1 | ||
|
|
||
| def xtest_mpinverse(self): | ||
| if not self.is_enabled(): | ||
| return | ||
|
|
||
| d = 1 / 16 | ||
| e = 1 / 16 | ||
| k = 2 / d | ||
| poly_deg = int((np.log(1 / e) / d)) | ||
| b = np.ceil(k * k * np.log(k / e)) | ||
| # odd polynomials for now | ||
| poly_deg = poly_deg if (np.mod(poly_deg, 2) == 1) else (poly_deg + 1) | ||
|
|
||
| # approximation to inverse function d/2x | ||
| def f(x): return np.where(x != 0, d / | ||
| 2 * (1 - (1 - x ** 2) ** b) / x, 0) | ||
|
|
||
| # The intput theta training values | ||
| th_in = np.arange(0, np.pi, np.pi / 30) | ||
| th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # The desired real part of p(x) which is the upper left value in the | ||
| # unitary of the qsp sequence | ||
| expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape[0])] | ||
| model = qsp_models.construct_qsp_model(poly_deg) | ||
| history = model.fit(x=th_in, y=expected_outputs, | ||
| epochs=5000, verbose=0) | ||
| # plot_loss(history) | ||
| # plot_qsp_response(f, model) | ||
| response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| model, return_all=True) | ||
| desired = f(np.cos(all_th)) | ||
| assert abs(response - desired).mean() < 0.1 | ||
|
|
||
| def test_ampamp1(self): | ||
| if not self.is_enabled(): | ||
| return | ||
| poly_deg = 19 | ||
|
|
||
| # approximation to inverse function d/2x | ||
| def f(x): return np.where(x < 0, -1, np.where(x > 0, 1, 0)) | ||
|
|
||
| # The intput theta training values | ||
| th_in = np.arange(0, np.pi, np.pi / 30) | ||
| th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # The desired real part of p(x) which is the upper left value in the | ||
| # unitary of the qsp sequence | ||
| expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape[0])] | ||
|
|
||
| model = qsp_models.construct_qsp_model(poly_deg) | ||
| history = model.fit(x=th_in, y=expected_outputs, | ||
| epochs=5000, verbose=0) | ||
| response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| model, return_all=True) | ||
| desired = f(np.cos(all_th)) | ||
| assert abs(response - desired).mean() < 0.1 | ||
| # import os | ||
| # import unittest | ||
|
|
||
| # import numpy as np | ||
| # # TODO: Currently silenced as transitioning away from Tensorflow. # import tensorflow as tf | ||
|
|
||
| # # TODO: Currently silenced as transitioning away from Tensorflow. # from pyqsp import qsp_models | ||
|
|
||
| # class Test_qsp_models(unittest.TestCase): | ||
| # ''' | ||
| # Unit tests for qsp_models component of pyqsp | ||
| # ''' | ||
| # def setUp(self): | ||
| # ''' | ||
| # Do imports here, so that this module can be entirely optional | ||
| # ''' | ||
| # if self.is_enabled(): | ||
| # pass | ||
| # else: | ||
| # print( | ||
| # "[pyqsp.test] Skipping qsp_model tests: export PYQSP_TEST_QSP_MODELS=1 to enable these tests") | ||
|
|
||
| # def is_enabled(self): | ||
| # enabled = 'PYQSP_TEST_QSP_MODELS' in os.environ | ||
| # return enabled | ||
|
|
||
| # def xtest_hsim1(self): | ||
| # if not self.is_enabled(): | ||
| # return | ||
| # t = 3 | ||
| # poly_deg = 6 | ||
| # def f(x): return np.cos(t * x) | ||
| # model = qsp_models.construct_qsp_model(poly_deg) | ||
|
|
||
| # # The intput theta training values | ||
| # th_in = np.arange(0, np.pi, np.pi / 50) | ||
| # th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # # The desired real part of p(x) which is the upper left value in the unitary of the qsp sequence | ||
| # # We also want that Re[q(x)] = 0 | ||
| # expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape)] | ||
|
|
||
| # model = qsp_models.construct_qsp_model(poly_deg) | ||
| # history = model.fit(x=th_in, y=expected_outputs, | ||
| # epochs=1000, verbose=0) | ||
| # # plot_loss(history) | ||
| # # plot_qsp_response(f, model) | ||
| # response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| # model, return_all=True) | ||
| # desired = f(np.cos(all_th)) | ||
| # assert abs(response - desired).mean() < 0.1 | ||
|
|
||
| # def xtest_mpinverse(self): | ||
| # if not self.is_enabled(): | ||
| # return | ||
|
|
||
| # d = 1 / 16 | ||
| # e = 1 / 16 | ||
| # k = 2 / d | ||
| # poly_deg = int((np.log(1 / e) / d)) | ||
| # b = np.ceil(k * k * np.log(k / e)) | ||
| # # odd polynomials for now | ||
| # poly_deg = poly_deg if (np.mod(poly_deg, 2) == 1) else (poly_deg + 1) | ||
|
|
||
| # # approximation to inverse function d/2x | ||
| # def f(x): return np.where(x != 0, d / | ||
| # 2 * (1 - (1 - x ** 2) ** b) / x, 0) | ||
|
|
||
| # # The intput theta training values | ||
| # th_in = np.arange(0, np.pi, np.pi / 30) | ||
| # th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # # The desired real part of p(x) which is the upper left value in the | ||
| # # unitary of the qsp sequence | ||
| # expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape[0])] | ||
| # model = qsp_models.construct_qsp_model(poly_deg) | ||
| # history = model.fit(x=th_in, y=expected_outputs, | ||
| # epochs=5000, verbose=0) | ||
| # # plot_loss(history) | ||
| # # plot_qsp_response(f, model) | ||
| # response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| # model, return_all=True) | ||
| # desired = f(np.cos(all_th)) | ||
| # assert abs(response - desired).mean() < 0.1 | ||
|
|
||
| # def test_ampamp1(self): | ||
| # if not self.is_enabled(): | ||
| # return | ||
| # poly_deg = 19 | ||
|
|
||
| # # approximation to inverse function d/2x | ||
| # def f(x): return np.where(x < 0, -1, np.where(x > 0, 1, 0)) | ||
|
|
||
| # # The intput theta training values | ||
| # th_in = np.arange(0, np.pi, np.pi / 30) | ||
| # th_in = tf.reshape(th_in, (th_in.shape[0], 1)) | ||
|
|
||
| # # The desired real part of p(x) which is the upper left value in the | ||
| # # unitary of the qsp sequence | ||
| # expected_outputs = [f(np.cos(th_in)), np.zeros(th_in.shape[0])] | ||
|
|
||
| # model = qsp_models.construct_qsp_model(poly_deg) | ||
| # history = model.fit(x=th_in, y=expected_outputs, | ||
| # epochs=5000, verbose=0) | ||
| # response, all_th, circuit_px, circuit_qx = qsp_models.compute_qsp_response( | ||
| # model, return_all=True) | ||
| # desired = f(np.cos(all_th)) | ||
| # assert abs(response - desired).mean() < 0.1 |