How to use converged method in molecule

Best Python code snippet using molecule_python

corrmut_test.py

Source:corrmut_test.py

...65    ###########66    def test_hard_conv_1(self):67        labels = [1] * 1068        pre_labels = [1] * 1069        converged = corrmut.has_converged(labels, pre_labels, "hard")70        assert converged71    def test_hard_conv_2(self):72        labels = [0] * 1073        pre_labels = [0] * 1074        converged = corrmut.has_converged(labels, pre_labels, "hard")75        assert converged76    def test_hard_conv_3(self):77        labels = [1] * 1078        pre_labels = [0] * 1079        converged = corrmut.has_converged(labels, pre_labels, "hard")80        assert not converged81    def test_hard_conv_4(self):82        labels = [0] * 1083        pre_labels = [1] * 1084        converged = corrmut.has_converged(labels, pre_labels, "hard")85        assert not converged86    def test_hard_conv_5(self):87        labels = [1, 1, 1, 0, 0, 0]88        pre_labels = [0, 0, 0, 1, 1, 1]89        converged = corrmut.has_converged(labels, pre_labels, "hard")90        assert not converged91    def test_hard_conv_6(self):92        labels = [1, 1, 1, 1, 0]93        pre_labels = [1, 1, 1, 1, 1]94        converged = corrmut.has_converged(labels, pre_labels, "hard")95        assert not converged96    def test_hard_conv_7(self):97        labels = [1, 1, 1, 0, 0, 0]98        pre_labels = [1, 1, 1, 0, 0, 0]99        converged = corrmut.has_converged(labels, pre_labels, "hard")100        assert converged101    ###########102    # Soft EM #103    ###########104    def test_soft_conv_1(self):105        labels = [1] * 10106        pre_labels = [1] * 10107        converged = corrmut.has_converged(labels, pre_labels, "soft")108        assert converged109    def test_soft_conv_2(self):110        labels = [0] * 10111        pre_labels = [0] * 10112        converged = corrmut.has_converged(labels, pre_labels, "soft")113        assert converged114    def test_soft_conv_3(self):115        labels = [1] * 10116        pre_labels = [0] * 10117        converged = corrmut.has_converged(labels, pre_labels, "soft")118        assert not converged119    def test_soft_conv_4(self):120        labels = [0] * 10121        pre_labels = [1] * 10122        converged = corrmut.has_converged(labels, pre_labels, "soft")123        assert not converged124    def test_soft_conv_5(self):125        labels = [1, 1, 1, 0, 0, 0]126        pre_labels = [0, 0, 0, 1, 1, 1]127        converged = corrmut.has_converged(labels, pre_labels, "soft")128        assert not converged129    def test_soft_conv_6(self):130        labels = [1, 1, 1, 1, 0]131        pre_labels = [1, 1, 1, 1, 1]132        converged = corrmut.has_converged(labels, pre_labels, "soft")133        assert not converged134    def test_soft_conv_7(self):135        labels = [1, 1, 1, 0, 0, 0]136        pre_labels = [1, 1, 1, 0, 0, 0]137        converged = corrmut.has_converged(labels, pre_labels, "soft")138        assert converged139    def test_soft_conv_8(self):140        labels = [0.99, 0.95, 0.90, 0.95, 0.1, 0.1, 0.001, 1e-3]141        pre_labels = [[0.99, 0.95, 0.90, 0.95, 0.1, 0.1, 0.001, 1e-16]]142        tol = 5e-3143        converged = corrmut.has_converged(labels, pre_labels, "soft", tol=tol)144        assert converged145    def test_soft_conv_9(self):146        labels = [0.99, 0.95, 0.90, 0.95, 0.1, 0.1, 0.001, 1e-3]147        pre_labels = [[0.99, 0.95, 0.90, 0.95, 0.1, 0.1, 0.001, 1e-32]]148        tol = 5e-3149        converged = corrmut.has_converged(labels, pre_labels, "soft", tol=tol)150        assert converged151    def test_soft_conv_10(self):152        # Difference between two arrays equals tolerance,153        # (would not be enough to reach convergence by itself)154        # but amount of observations must also be taken into account!155        labels = [0.99, 0.99, 0.1, 0.1, 0.005]156        pre_labels = [0.99, 0.99, 0.1, 0.1, 0.01]157        tol = 5e-3158        converged = corrmut.has_converged(labels, pre_labels, "soft", tol=tol)159        assert converged160#####################161# Alternative model #162#####################163class TestGetPosteriorLogProbs():164    """165    Class to test the corrmut.get_posterior_logprobs function166    """167    def test_logprobs_1(self):168        # Tested using sklearn 0.19.2169        # Covariation between two toy alignments170        Y = [3, 11, 3, 11, 3, 11]171        X = np.array([[11, 3, 11, 3, 11, 3]])172        exp_logprobs = np.log([0.79807613, 0.79822048, 0.79807613,...

plot_vary_multipliers_v2_data.py

Source:plot_vary_multipliers_v2_data.py

1"""2Plot v2 of Fig. 4's parameter sweep, which varies over overlap length, finger width, initial front gap, and support spring width3"""4import os5file_location = os.path.abspath(os.path.dirname( __file__))6dir_location = os.path.abspath(os.path.join(file_location, '..'))7import sys8sys.path.append(file_location)9sys.path.append(dir_location)10import numpy as np11import matplotlib.pyplot as plt12from scipy.io import loadmat, savemat13from datetime import datetime14plt.rc('font', size=11.5)15plt.rcParams['legend.fontsize'] = 11.516def setup_plot(len_x, len_y, plt_title=None, x_label="", y_label=""):17    fig, axs = plt.subplots(len_x, len_y)18    if plt_title is not None:19        fig.suptitle(plt_title)20    if x_label or y_label:21        # add a big axis, hide frame22        fig.add_subplot(111, frameon=False)23        # hide tick and tick label of the big axis24        plt.tick_params(labelcolor='none', which='both', top=False, bottom=False, left=False, right=False)25        plt.xlabel(x_label, fontsize=12)26        plt.ylabel(y_label, fontsize=12)27    return fig, axs28def display_stats(x_converged, times_converged, label, nom):29    # Get helpful stats for paper30    # print(label, x_converged, times_converged)31    try:32        time_50_up = times_converged[np.where(np.isclose(x_converged, nom*1.25e6))[0][0]]33        time_nominal = times_converged[np.where(np.isclose(x_converged, nom*1.e6))[0][0]]34        time_50_down = times_converged[np.where(np.isclose(x_converged, nom*0.8e6))[0][0]]35        print("{}: con=0.75: {} (Ratio: {}), con=1: {}, con=1.25: {} (Ratio: {})".format(36            label, time_50_down, 1 - time_50_down/time_nominal, time_nominal, time_50_up, 1 - time_50_up/time_nominal37        ))38    except Exception as e:39        print(e)40now = datetime.now()41name_clarifier = "_20211024_01_37_55_vary_multipliers_undercut=0.400_Fes=v2_Fb=v2_modified"42timestamp = now.strftime("%Y%m%d_%H_%M_%S") + name_clarifier43print(timestamp)44filename = "../data/20211024_01_37_55_vary_multipliers_undercut=0.400_Fes=v2_Fb=v2.npy"45fileData = np.load(filename, allow_pickle=True)46process, fingerLcon_range, fingerWcon_range, gfcon_range, supportWcon_range, fingerLnom, fingerWnom, gfnom, supportWnom, data, fig = fileData47plt.close()48fig, axs = setup_plot(2, 2, y_label=r"Time ($\mu$s)")49### fingerL50x_converged, times_converged = data["fingerLpullin"]51x_converged = np.array(x_converged)52times_converged = np.array(times_converged)53idx = np.where(x_converged < 95)54x_converged = x_converged[idx]55times_converged = times_converged[idx]56axs[0, 0].plot(x_converged, times_converged, 'b')57display_stats(x_converged, times_converged, "Pullin fingerLcon", fingerLnom)58x_converged, times_converged = data["fingerLrelease"]59x_converged = np.array(x_converged)60times_converged = np.array(times_converged)61idx = np.where(x_converged < 95)62x_converged = x_converged[idx]63times_converged = times_converged[idx]64axs[0, 0].plot(x_converged, times_converged, 'r')65axs[0, 0].set_title(r"Varying $L_{ol}$", fontsize=12)66axs[0, 0].axvline(fingerLnom*1e6, color='k', linestyle='--')67display_stats(x_converged, times_converged, "Release fingerLcon", fingerLnom)68label = r"$L_{ol}=$" + "{:0.1f}".format(fingerLnom*1e6) + r' $\mu$m'69axs[0, 0].annotate(label, xy=(0.73, 0.96), xycoords='axes fraction', color='k',70                   xytext=(0, 0), textcoords='offset points', ha='right', va='top')71axs[0, 0].set_xlabel(r'$L_{ol}  (\mu$m)')72### fingerW73x_converged, times_converged = data["fingerWpullin"]74x_converged = np.array(x_converged)75times_converged = np.array(times_converged)76# idx = np.where(x_converged > 3.425)77idx = np.where(np.array(x_converged) > 3.5)78# idx = np.where(x_converged >= 4.)79x_converged = x_converged[idx]80times_converged = times_converged[idx]81axs[0, 1].plot(x_converged, times_converged, 'b')82display_stats(x_converged, times_converged, "Pullin fingerWcon", fingerWnom)83x_converged, times_converged = data["fingerWrelease"]84x_converged = np.array(x_converged)85times_converged = np.array(times_converged)86# idx = np.where(np.array(x_converged) > 3.425)87idx = np.where(np.array(x_converged) > 3.5)88# idx = np.where(x_converged >= 4.)89x_converged = x_converged[idx]90times_converged = times_converged[idx]91axs[0, 1].plot(x_converged, times_converged, 'r')92axs[0, 1].set_title(r"Varying $w_f$", fontsize=12)93axs[0, 1].axvline(fingerWnom*1e6, color='k', linestyle='--')94display_stats(x_converged, times_converged, "Release fingerWcon", fingerWnom)95label = r"$w_f=$" + "{:0.1f}".format(fingerWnom*1e6) + r' $\mu$m'96axs[0, 1].annotate(label, xy=(0.29, 0.6), xycoords='axes fraction', color='k',97                   xytext=(0, 0), textcoords='offset points', ha='left', va='top')98axs[0, 1].set_xlabel(r'$w_f  (\mu$m)')99### gf0100x_converged, times_converged = data["gfpullin"]101axs[1, 0].plot(x_converged, times_converged, 'b')102display_stats(x_converged, times_converged, "Pullin gf0con", gfnom)103max_x = max(x_converged)104x_converged, times_converged = data["gfrelease"]105x_converged = np.array(x_converged)106times_converged = np.array(times_converged)107idx = np.where(np.array(x_converged) < max_x)108x_converged = x_converged[idx]109times_converged = times_converged[idx]110axs[1, 0].plot(x_converged, times_converged, 'r')111axs[1, 0].set_title(r"Varying $x_0$", fontsize=12)112axs[1, 0].axvline(gfnom*1e6, color='k', linestyle='--')113display_stats(x_converged, times_converged, "Release gf0con", gfnom)114label = r"$x_0=$" + "\n" + "{:0.2f}".format(gfnom*1e6) + r' $\mu$m'115axs[1, 0].annotate(label, xy=(0.57, 0.96), xycoords='axes fraction', color='k',116                   xytext=(0, 0), textcoords='offset points', ha='left', va='top')117axs[1, 0].set_xlabel(r'$x_0  (\mu$m)')118### supportW119x_converged, times_converged = data["supportWpullin"]120axs[1, 1].plot(x_converged, times_converged, 'b')121display_stats(x_converged, times_converged, "Pullin supportWcon", supportWnom)122x_converged, times_converged = data["supportWrelease"]123axs[1, 1].plot(x_converged, times_converged, 'r')124axs[1, 1].set_title(r"Varying $w_{spr}$", fontsize=12)125axs[1, 1].axvline(supportWnom*1e6, color='k', linestyle='--')126display_stats(x_converged, times_converged, "Release supportWcon", supportWnom)127label = r"$w_{spr}=$" + "{:0.1f}".format(supportWnom*1e6) + r' $\mu$m'128axs[1, 1].annotate(label, xy=(0.47, 0.96), xycoords='axes fraction', color='k',129                   xytext=(0, 0), textcoords='offset points', ha='left', va='top')130axs[1, 1].set_xlabel(r'$w_{spr}  (\mu$m)')131axs[1, 1].legend(["Pull-in", "Release"], loc='center right')132plt.tight_layout()133timestamp = "20211024_01_37_55_vary_multipliers_undercut=0.400_Fes=v2_Fb=v2_modified_v4"134# plt.savefig("../figures/" + timestamp + ".png")135# plt.savefig("../figures/" + timestamp + ".pdf")...

library.py

Source:library.py

1'''2    Copyright (C) 2020 Anne Hartebrodt3    This program is free software; you can redistribute it and/or modify4    it under the terms of the GNU General Public License as published by5    the Free Software Foundation; either version 2 of the License, or6    (at your option) any later version.7    This program is distributed in the hope that it will be useful,8    but WITHOUT ANY WARRANTY; without even the implied warranty of9    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the10    GNU General Public License for more details.11    You should have received a copy of the GNU General Public License along12    with this program; if not, write to the Free Software Foundation, Inc.,13    51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.14    Authors: Anne Hartebrodt15'''16import numpy as np17import scipy as sc18import svd.shared_functions as sh19import scipy.linalg as la20import svd.comparison as co21def local1(data, G_i):22    return np.dot(data, G_i)23def local2(data, H_pooled, G_previous):24    return G_previous + np.dot(data.T, H_pooled)25def pooling(H_all):26    H_pooled = H_all[0]27    for h in H_all[1:len(H_all)]:28        H_pooled = H_pooled + h29    #H_pooled = H_pooled / len(H_all)30    H_pooled = H_pooled / np.linalg.norm(H_pooled)31    return H_pooled32def standalone(data, k=10):33    G_i = sh.generate_random_gaussian(data.shape[1], k) # phi134    G_i, Q = la.qr(G_i, mode='economic')35    converged = False36    previous = G_i37    previous_h = sh.generate_random_gaussian(data.shape[0], k)38    iterations = 039    while not converged:40        iterations = iterations+141        print(iterations)42        H_i = np.dot(data, G_i) # YiPhii , gamma if standalone43        G_i =  np.dot(data.T, H_i)  + previous44        G_i, Q = la.qr(G_i, mode='economic') # 'compute residuals'45        converged, sum = convergence_checker(H_i, previous_h, epsilon=0.0000001)46        previous_h = H_i47        previous = G_i48    return G_i49def convergence_checker_d(current, previous, epsilon=0.000001):50    '''51    Checks the convergence52    Args:53        H_i:54        N_list:55        previous:56        epsilon:57    Returns: -1 if convergence not reached58    '''59    sum = 060    for i in range(current.shape[1]):61        ra = np.dot(current[:, i].T, current[:, i]) / sc.linalg.norm(current[:, i])62        sum = np.abs(ra)63    if np.abs(sum-previous) < epsilon:64        return -165    else:66        return ra67def convergence_checker_angle(current, previous, epsilon=0.000001, return_converged=False):68    """69    Check if angle between consecutive eigenvectors is nearing 070    Args:71        current: Current eigenvalue guesss72        previous: Previous eigenvalue guess73        epsilon: tolerance74        return_converged: verbosity75    Returns:76    """77    sum = 078    converged = True79    converged_eigenvectors = []80    deltas = []81    for i in range(current.shape[1]):82        ra = np.dot(current[:,i].T, previous[:,i])83        if ra >= 1-epsilon:84            converged_eigenvectors.append(i)85        else:86            # if one of the eigenvalues has not converged set overall convergence to true87            converged = False88    if return_converged:89        return converged, sum,  converged_eigenvectors, deltas90    else:91        return converged, sum92def convergence_checker(current, previous, epsilon=0.000001, return_converged=False):93    """94    Convergence checked via Raleigh coefficient.95    Args:96        current:97        previous:98        epsilon:99        return_converged:100    Returns:101    """102    sum = 0103    converged = True104    converged_eigenvectors = []105    deltas = []106    for i in range(current.shape[1]):107        ra = np.dot(current[:,i].T, current[:,i])/sc.linalg.norm(current[:,i])108        rap = np.dot(previous[:,i].T, previous[:,i])/sc.linalg.norm(previous[:,i])109        sum = sum + np.abs(ra-rap)110        deltas.append(np.abs(ra-rap))111        if np.abs(ra-rap) > epsilon:112            converged=False113        else:114            converged_eigenvectors.append(i)115    if return_converged:116        return converged, sum,  converged_eigenvectors, deltas117    else:118        return converged, sum119def convergence_checker_a(current, previous,alpha_norm, alpha_norm_prev, epsilon=0.000001, return_converged=False):120    """121    weird convergence criterion emplyed by Guo et al.122    Args:123        current: Current H124        previous: previous H125        alpha_norm: Current G norm126        alpha_norm_prev: previous G norm127        epsilon: tolerance128        return_converged: Which values to return129    Returns:130    """131    sum = 0132    converged = True133    converged_eigenvectors = []134    deltas = []135    for i in range(current.shape[1]):136        ra = np.dot(current[:,i].T, current[:,i])/alpha_norm[i]137        rap = np.dot(previous[:,i].T, previous[:,i])/alpha_norm_prev[i]138        sum = sum + np.abs(ra-rap)139        deltas.append(np.abs(ra-rap))140        if np.abs(ra-rap) > epsilon:141            converged=False142        else:143            converged_eigenvectors.append(i)144    if return_converged:145        return converged, sum,  converged_eigenvectors, deltas146    else:147        return converged, sum148def standalone2(data, first):149    G_i = sh.generate_random_gaussian(data.shape[1], 1)150    G_i = G_i - np.dot(np.inner(G_i, first.T) , first.T)151    print(np.asarray(G_i).T)152    print(first)153    print(co.angle(np.asarray(G_i).T, np.asarray(first).T))154    Q, R = la.qr(np.asarray(np.concatenate([ first.T,G_i], axis = 1)))155    print(Q[:,0])156    G_i = Q[:,1]157    converged = False158    previous = G_i159    previous_h = sh.generate_random_gaussian(data.shape[0], 1)160    iterations = 0161    while not converged:162        iterations = iterations+1163        print(iterations)164        H_i = np.dot(data, G_i) # YiPhii , gamma if standalone165        G_i =  np.dot(data.T, H_i)  + previous166        Q, R = la.qr(np.asarray(np.stack([first.flatten(),G_i.T])).T)167        G_i = Q[:, 1]168        converged = convergence_checker(H_i, previous_h)169        previous_h = H_i170        previous = G_i171    return G_i172def get_initial_eigenvector_k(V):173    '''174     ap = a -sum over k-1 <a, ai>ai175    Args:176        V:177    Returns:178    '''179    a = sh.generate_random_gaussian(1, V.shape[0])180    sum = np.zeros(V.shape[0])181    for v in range(V.shape[1]):182        sum = sum + np.dot(np.dot(a, v), v)183    ap = a - sum...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.