from Ccoder import Ccoder import unittest import copy import json import os import shutil class TestCcoder(unittest.TestCase): def setUp(self): dpkgRoot = os.path.join('..' ,'examples', 'noise') dpkg = json.load(open(os.path.join(dpkgRoot, 'ssm.json'))) self.m_noise = Ccoder(dpkgRoot, copy.deepcopy(dpkg)) m_diff = copy.deepcopy(dpkg) del m_diff['reactions'][2]['white_noise'] del m_diff['reactions'][3]['white_noise'] m_diff['sde'] = { 'drift': [ {'name': 'r0_paris', 'f': 0.0, 'transformation': 'log(r0_paris)'}, {'name': 'r0_nyc', 'f': 0.0, 'transformation': 'log(r0_nyc)'} ], 'dispertion': [['vol',0],[0,'vol']] } self.m_diff = Ccoder(dpkgRoot, m_diff) m_noise2 = copy.deepcopy(dpkg) del m_noise2['reactions'][2]['white_noise'] del m_noise2['reactions'][3]['white_noise'] m_noise2['reactions'][0]['white_noise'] = {'name':'noise_SI', 'sd': 'sto'} m_noise2['reactions'][1]['white_noise'] = {'name':'noise_SI2', 'sd': 'sto'} self.m_noise2 = Ccoder(dpkgRoot, m_noise2) m_noise3 = copy.deepcopy(dpkg) del m_noise3['reactions'][2]['white_noise'] del m_noise3['reactions'][3]['white_noise'] m_noise3['reactions'][4]['white_noise'] = {'name':'noise_SI', 'sd': 'sto'} m_noise3['reactions'][5]['white_noise'] = {'name':'noise_SI2', 'sd': 'sto'} self.m_noise3 = Ccoder(dpkgRoot, m_noise3) m_noise4 = copy.deepcopy(dpkg) del m_noise4['reactions'][2]['white_noise'] del m_noise4['reactions'][3]['white_noise'] m_noise4['reactions'][8]['white_noise'] = {'name':'noise_SI', 'sd': 'sto'} m_noise4['reactions'][9]['white_noise'] = {'name':'noise_SI2', 'sd': 'sto'} self.m_noise4 = Ccoder(dpkgRoot, m_noise4) m_noise5 = copy.deepcopy(dpkg) del m_noise5['reactions'][2]['white_noise'] del m_noise5['reactions'][3]['white_noise'] m_noise5['reactions'][10]['white_noise'] = {'name':'noise_SI', 'sd': 'sto'} m_noise5['reactions'][11]['white_noise'] = {'name':'noise_SI2', 'sd': 'sto'} self.m_noise5 = Ccoder(dpkgRoot, m_noise5) m_noise6 = copy.deepcopy(dpkg) m_noise6['reactions'][4]['white_noise'] = {'name':'noise_SI', 'sd': 'sto'} m_noise6['reactions'][5]['white_noise'] = {'name':'noise_SI2', 'sd': 'sto'} self.m_noise6 = Ccoder(dpkgRoot, m_noise6) m_noise7 = copy.deepcopy(dpkg) m_noise7['reactions'][4]['white_noise'] = {'name':'noise_SI23', 'sd': 'sto'} m_noise7['reactions'][5]['white_noise'] = {'name':'noise_SI24', 'sd': 'sto'} self.m_noise7 = Ccoder(dpkgRoot, m_noise7) m_diff2 = copy.deepcopy(dpkg) del m_diff2['reactions'][2]['white_noise'] del m_diff2['reactions'][3]['white_noise'] m_diff2['sde'] = copy.deepcopy(m_diff['sde']) m_diff2['reactions'].append({'from': 'R_paris', 'to': 'I_paris', 'rate': 'correct_rate(v)', 'description':'testing'}) m_diff2['reactions'].append({'from': 'R_nyc', 'to': 'I_nyc', 'rate': 'correct_rate(v)', 'description':'testing'}) self.m_diff2 = Ccoder(dpkgRoot, m_diff2) def test_eval_Q(self): calc_Q = self.m_noise.eval_Q() # testing env sto only # order: I_nyc, I_paris, S_nyc, S_paris, inc_all , inc_nyc term_paris = '((((r0_paris/N_paris*v*I_paris)*S_paris)*((sto)**2))*((r0_paris/N_paris*v*I_paris)*S_paris)))' term_nyc = '((((r0_nyc/N_nyc*v*I_nyc)*S_nyc)*((sto)**2))*((r0_nyc/N_nyc*v*I_nyc)*S_nyc)))' Q_cm = calc_Q["no_dem_sto"]["Q_cm"] self.assertEqual(Q_cm[0][0], '((1)*'+term_nyc+'*(1)') self.assertEqual(Q_cm[0][1], 0) self.assertEqual(Q_cm[0][2], '((1)*'+term_nyc+'*(-1)') self.assertEqual(Q_cm[0][3], 0) self.assertEqual(Q_cm[0][4], 0) self.assertEqual(Q_cm[0][5], '((1)*'+term_nyc+'*(1)') self.assertEqual(Q_cm[1][0], 0) self.assertEqual(Q_cm[1][1], '((1)*'+term_paris+'*(1)') self.assertEqual(Q_cm[1][2], 0) self.assertEqual(Q_cm[1][3], '((1)*'+term_paris+'*(-1)') self.assertEqual(Q_cm[1][4], 0) self.assertEqual(Q_cm[1][5], 0) self.assertEqual(Q_cm[2][0], '((-1)*'+term_nyc+'*(1)') self.assertEqual(Q_cm[2][1], 0) self.assertEqual(Q_cm[2][2], '((-1)*'+term_nyc+'*(-1)') self.assertEqual(Q_cm[2][3], 0) self.assertEqual(Q_cm[2][4], 0) self.assertEqual(Q_cm[2][5], '((-1)*'+term_nyc+'*(1)') self.assertEqual(Q_cm[3][0], 0) self.assertEqual(Q_cm[3][1], '((-1)*'+term_paris+'*(1)') self.assertEqual(Q_cm[3][2], 0) self.assertEqual(Q_cm[3][3], '((-1)*'+term_paris+'*(-1)') self.assertEqual(Q_cm[3][4], 0) self.assertEqual(Q_cm[3][5], 0) self.assertEqual(Q_cm[4][0], 0) self.assertEqual(Q_cm[4][1], 0) self.assertEqual(Q_cm[4][2], 0) self.assertEqual(Q_cm[4][3], 0) self.assertEqual(Q_cm[4][4], 0) self.assertEqual(Q_cm[4][5], 0) self.assertEqual(Q_cm[5][0], '((1)*'+term_nyc+'*(1)') self.assertEqual(Q_cm[5][1], 0) self.assertEqual(Q_cm[5][2], '((1)*'+term_nyc+'*(-1)') self.assertEqual(Q_cm[5][3], 0) self.assertEqual(Q_cm[5][4], 0) self.assertEqual(Q_cm[5][5], '((1)*'+term_nyc+'*(1)') # testing dem sto only term1_p = '(mu_b_paris*N_paris))' term2_p = '((r0_paris/N_paris*v*I_paris)*S_paris))' term3_p = '((correct_rate(v))*I_paris))' term4_p = '((mu_d_paris)*S_paris))' term5_p = '((mu_d_paris)*I_paris))' term1_n = '(mu_b_nyc*N_nyc))' term2_n = '((r0_nyc/N_nyc*v*I_nyc)*S_nyc))' term3_n = '((correct_rate(v))*I_nyc))' term4_n = '((mu_d_nyc)*S_nyc))' term5_n = '((mu_d_nyc)*I_nyc))' # order: I_nyc, I_paris, S_nyc, S_paris, inc_all , inc_nyc Q_cm = calc_Q["no_env_sto"]["Q_cm"] self.assertEqual(Q_cm[0][0], '((1)*'+term2_n+'*(1) + ((-1)*'+term3_n+'*(-1) + ((-1)*'+term5_n+'*(-1)') self.assertEqual(Q_cm[0][1], 0) self.assertEqual(Q_cm[0][2], '((1)*'+term2_n+'*(-1)') self.assertEqual(Q_cm[0][3], 0) self.assertEqual(Q_cm[0][4], '((-1)*'+term3_n+'*(1) + ((-1)*'+term5_n+'*(1)') self.assertEqual(Q_cm[0][5], '((1)*'+term2_n+'*(1)') self.assertEqual(Q_cm[1][0], 0) self.assertEqual(Q_cm[1][1], '((1)*'+term2_p+'*(1) + ((-1)*'+term3_p+'*(-1) + ((-1)*'+term5_p+'*(-1)') self.assertEqual(Q_cm[1][2], 0) self.assertEqual(Q_cm[1][3], '((1)*'+term2_p+'*(-1)') self.assertEqual(Q_cm[1][4], '((-1)*'+term3_p+'*(1) + ((-1)*'+term5_p+'*(1)') self.assertEqual(Q_cm[1][5], 0) self.assertEqual(Q_cm[2][0], '((-1)*'+term2_n+'*(1)') self.assertEqual(Q_cm[2][1], 0) self.assertEqual(Q_cm[2][2], '((1)*'+term1_n+'*(1) + ((-1)*'+term2_n+'*(-1) + ((-1)*'+term4_n+'*(-1)') self.assertEqual(Q_cm[2][3], 0) self.assertEqual(Q_cm[2][4], 0) self.assertEqual(Q_cm[2][5], '((-1)*'+term2_n+'*(1)') self.assertEqual(Q_cm[3][0], 0) self.assertEqual(Q_cm[3][1], '((-1)*'+term2_p+'*(1)') self.assertEqual(Q_cm[3][2], 0) self.assertEqual(Q_cm[3][3], '((1)*'+term1_p+'*(1) + ((-1)*'+term2_p+'*(-1) + ((-1)*'+term4_p+'*(-1)') self.assertEqual(Q_cm[3][4], 0) self.assertEqual(Q_cm[3][5], 0) self.assertEqual(Q_cm[4][0], '((1)*'+term3_n+'*(-1) + ((1)*'+term5_n+'*(-1)') self.assertEqual(Q_cm[4][1], '((1)*'+term3_p+'*(-1) + ((1)*'+term5_p+'*(-1)') self.assertEqual(Q_cm[4][2], 0) self.assertEqual(Q_cm[4][3], 0) self.assertEqual(Q_cm[4][4], '((1)*'+term3_p+'*(1) + ((1)*'+term3_n+'*(1) + ((1)*'+term5_p+'*(1) + ((1)*'+term5_n+'*(1)') self.assertEqual(Q_cm[4][5], 0) self.assertEqual(Q_cm[5][0], '((1)*'+term2_n+'*(1)') self.assertEqual(Q_cm[5][1], 0) self.assertEqual(Q_cm[5][2], '((1)*'+term2_n+'*(-1)') self.assertEqual(Q_cm[5][3], 0) self.assertEqual(Q_cm[5][4], 0) self.assertEqual(Q_cm[5][5], '((1)*'+term2_n+'*(1)') def test_eval_Q_tricky_cases(self): calc_Q = self.m_noise2.eval_Q() # testing env sto only for m_noise2 : WN on U->S term_p = '(((mu_b_paris*N_paris)*((sto)**2))*(mu_b_paris*N_paris)))' term_n = '(((mu_b_nyc*N_nyc)*((sto)**2))*(mu_b_nyc*N_nyc)))' for i in range(5): for j in range(5): if i==2 and j == 2: self.assertEqual(calc_Q["no_dem_sto"]["Q_cm"][i][j],'((1)*'+term_n+'*(1)') elif i==3 and j == 3: self.assertEqual(calc_Q["no_dem_sto"]["Q_cm"][i][j],'((1)*'+term_p+'*(1)') else: self.assertEqual(calc_Q["no_dem_sto"]["Q_cm"][i][j],0) calc_Q = self.m_noise3.eval_Q() # testing env sto only for m_noise3 : WN on I->R term_p = '((((correct_rate(v))*I_paris)*((sto)**2))*((correct_rate(v))*I_paris)))' term_n = '((((correct_rate(v))*I_nyc)*((sto)**2))*((correct_rate(v))*I_nyc)))' Q_cm = calc_Q["no_dem_sto"]["Q_cm"] for i in range(5): for j in range(5): if i==0 and j == 0: self.assertEqual(Q_cm[i][j], '((-1)*'+term_n+'*(-1)') elif i==1 and j == 1: self.assertEqual(Q_cm[i][j], '((-1)*'+term_p+'*(-1)') elif i==0 and j == 4: self.assertEqual(Q_cm[i][j], '((-1)*'+term_n+'*(1)') elif i==4 and j == 0: self.assertEqual(Q_cm[i][j], '((1)*'+term_n+'*(-1)') elif i==1 and j == 4: self.assertEqual(Q_cm[i][j], '((-1)*'+term_p+'*(1)') elif i==4 and j == 1: self.assertEqual(Q_cm[i][j], '((1)*'+term_p+'*(-1)') elif i==4 and j == 4: self.assertEqual(Q_cm[i][j], '((1)*'+term_p+'*(1) + ((1)*'+term_n+'*(1)') else: self.assertEqual(Q_cm[i][j], 0) calc_Q = self.m_noise4.eval_Q() # testing env sto only for m_noise4 : WN on I->U term_p = '((((mu_d_paris)*I_paris)*((sto)**2))*((mu_d_paris)*I_paris)))' term_n = '((((mu_d_nyc)*I_nyc)*((sto)**2))*((mu_d_nyc)*I_nyc)))' Q_cm = calc_Q["no_dem_sto"]["Q_cm"] for i in range(5): for j in range(5): if i==0 and j == 0: self.assertEqual(Q_cm[i][j], '((-1)*'+term_n+'*(-1)') elif i==1 and j == 1: self.assertEqual(Q_cm[i][j], '((-1)*'+term_p+'*(-1)') elif i==1 and j == 4: self.assertEqual(Q_cm[i][j], '((-1)*'+term_p+'*(1)') elif i==4 and j == 1: self.assertEqual(Q_cm[i][j], '((1)*'+term_p+'*(-1)') elif i==0 and j == 4: self.assertEqual(Q_cm[i][j], '((-1)*'+term_n+'*(1)') elif i==4 and j == 0: self.assertEqual(Q_cm[i][j], '((1)*'+term_n+'*(-1)') elif i==4 and j == 4: self.assertEqual(Q_cm[i][j], '((1)*'+term_p+'*(1) + ((1)*'+term_n+'*(1)') else: self.assertEqual(Q_cm[i][j], 0) calc_Q = self.m_noise5.eval_Q() # testing env sto only for m_noise5 : WN on R->U for i in range(5): for j in range(5): self.assertEqual(calc_Q["no_dem_sto"]["Q_cm"][i][j],0) calc_Q = self.m_noise6.eval_Q() # testing env sto only for m_noise6 : correlated WN on I->R and S->I term1_p = '((r0_paris/N_paris*v*I_paris)*S_paris)' term2_p = '((correct_rate(v))*I_paris)' term1_n = '((r0_nyc/N_nyc*v*I_nyc)*S_nyc)' term2_n = '((correct_rate(v))*I_nyc)' Q_cm = calc_Q["no_dem_sto"]["Q_cm"] self.assertEqual(Q_cm[0][0], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+') + (-1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(1) + ((1)*(('+term1_n+'*((sto)**2))*'+term2_n+') + (-1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[0][1], 0) self.assertEqual(Q_cm[0][2], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+') + (-1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[0][3], 0) self.assertEqual(Q_cm[0][4], '((1)*(('+term1_n+'*((sto)**2))*'+term2_n+') + (-1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[0][5], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+') + (-1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[1][0], 0) self.assertEqual(Q_cm[1][1], '((1)*(('+term1_p+'*((sto)**2))*'+term1_p+') + (-1)*(('+term2_p+'*((sto)**2))*'+term1_p+'))*(1) + ((1)*(('+term1_p+'*((sto)**2))*'+term2_p+') + (-1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(-1)') self.assertEqual(Q_cm[1][2], 0) self.assertEqual(Q_cm[1][3], '((1)*(('+term1_p+'*((sto)**2))*'+term1_p+') + (-1)*(('+term2_p+'*((sto)**2))*'+term1_p+'))*(-1)') self.assertEqual(Q_cm[1][4], '((1)*(('+term1_p+'*((sto)**2))*'+term2_p+') + (-1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(1)') self.assertEqual(Q_cm[1][5], 0) self.assertEqual(Q_cm[2][0], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1) + ((-1)*(('+term1_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[2][1], 0) self.assertEqual(Q_cm[2][2], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[2][3], 0) self.assertEqual(Q_cm[2][4], '((-1)*(('+term1_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[2][5], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[3][0], 0) self.assertEqual(Q_cm[3][1], '((-1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(1) + ((-1)*(('+term1_p+'*((sto)**2))*'+term2_p+'))*(-1)') self.assertEqual(Q_cm[3][2], 0) self.assertEqual(Q_cm[3][3], '((-1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(-1)') self.assertEqual(Q_cm[3][4], '((-1)*(('+term1_p+'*((sto)**2))*'+term2_p+'))*(1)') self.assertEqual(Q_cm[3][5], 0) self.assertEqual(Q_cm[4][0], '((1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(1) + ((1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[4][1], '((1)*(('+term2_p+'*((sto)**2))*'+term1_p+'))*(1) + ((1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(-1)') self.assertEqual(Q_cm[4][2], '((1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[4][3], '((1)*(('+term2_p+'*((sto)**2))*'+term1_p+'))*(-1)') self.assertEqual(Q_cm[4][4], '((1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(1) + ((1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[4][5], '((1)*(('+term2_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[5][0], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1) + ((1)*(('+term1_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[5][1], 0) self.assertEqual(Q_cm[5][2], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[5][3], 0) self.assertEqual(Q_cm[5][4], '((1)*(('+term1_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[5][5], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') calc_Q = self.m_noise7.eval_Q() # testing env sto only for m_noise7 : uncorrelated WN on I->R and S->I term1_p = '((r0_paris/N_paris*v*I_paris)*S_paris)' term2_p = '((correct_rate(v))*I_paris)' term1_n = '((r0_nyc/N_nyc*v*I_nyc)*S_nyc)' term2_n = '((correct_rate(v))*I_nyc)' Q_cm = calc_Q["no_dem_sto"]["Q_cm"] self.assertEqual(Q_cm[0][0], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1) + ((-1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[0][1], 0) self.assertEqual(Q_cm[0][2], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[0][3], 0) self.assertEqual(Q_cm[0][4], '((-1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[0][5], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[1][0], 0) self.assertEqual(Q_cm[1][1], '((1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(1) + ((-1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(-1)') self.assertEqual(Q_cm[1][2], 0) self.assertEqual(Q_cm[1][3], '((1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(-1)') self.assertEqual(Q_cm[1][4], '((-1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(1)') self.assertEqual(Q_cm[1][5], 0) self.assertEqual(Q_cm[2][0], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[2][1], 0) self.assertEqual(Q_cm[2][2], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[2][3], 0) self.assertEqual(Q_cm[2][4], 0) self.assertEqual(Q_cm[2][5], '((-1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[3][0], 0) self.assertEqual(Q_cm[3][1], '((-1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(1)') self.assertEqual(Q_cm[3][2], 0) self.assertEqual(Q_cm[3][3], '((-1)*(('+term1_p+'*((sto)**2))*'+term1_p+'))*(-1)') self.assertEqual(Q_cm[3][4], 0) self.assertEqual(Q_cm[3][5], 0) self.assertEqual(Q_cm[4][0], '((1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(-1)') self.assertEqual(Q_cm[4][1], '((1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(-1)') self.assertEqual(Q_cm[4][2], 0) self.assertEqual(Q_cm[4][3], 0) self.assertEqual(Q_cm[4][4], '((1)*(('+term2_p+'*((sto)**2))*'+term2_p+'))*(1) + ((1)*(('+term2_n+'*((sto)**2))*'+term2_n+'))*(1)') self.assertEqual(Q_cm[4][5], 0) self.assertEqual(Q_cm[5][0], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') self.assertEqual(Q_cm[5][1], 0) self.assertEqual(Q_cm[5][2], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(-1)') self.assertEqual(Q_cm[5][3], 0) self.assertEqual(Q_cm[5][4], 0) self.assertEqual(Q_cm[5][5], '((1)*(('+term1_n+'*((sto)**2))*'+term1_n+'))*(1)') def test_jac(self): step_ode_sde = self.m_noise.step_ode_sde() jac = self.m_diff.jac(step_ode_sde['sf']) # testing jac # I ode - ((v)*I) - ((mu_d)*I) + ((r0/N*v*I)*S) self.assertEqual(jac['caches'][jac['jac'][0][0]], '-gsl_spline_eval(calc->spline[ORDER_mu_d_nyc],t,calc->acc[ORDER_mu_d_nyc])-(gsl_vector_get(par,ORDER_v))+X[ORDER_S_nyc]*diffed[ORDER_diff__r0_nyc]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_nyc],t,calc->acc[ORDER_N_nyc])') self.assertEqual(jac['caches'][jac['jac'][1][1]], '-gsl_spline_eval(calc->spline[ORDER_mu_d_paris],t,calc->acc[ORDER_mu_d_paris])-(gsl_vector_get(par,ORDER_v))+X[ORDER_S_paris]*diffed[ORDER_diff__r0_paris]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_paris],t,calc->acc[ORDER_N_paris])') self.assertEqual(jac['caches'][jac['jac_diff'][1][0]['value']], '0') self.assertEqual(jac['caches'][jac['jac_diff'][0][1]['value']], '0') # S ode - ((r0/N*v*I)*S) - ((mu_d)*S) + (mu_b*N) self.assertEqual(jac['caches'][jac['jac'][2][2]], '-X[ORDER_I_nyc]*diffed[ORDER_diff__r0_nyc]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_nyc],t,calc->acc[ORDER_N_nyc])-gsl_spline_eval(calc->spline[ORDER_mu_d_nyc],t,calc->acc[ORDER_mu_d_nyc])') self.assertEqual(jac['caches'][jac['jac'][3][3]], '-X[ORDER_I_paris]*diffed[ORDER_diff__r0_paris]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_paris],t,calc->acc[ORDER_N_paris])-gsl_spline_eval(calc->spline[ORDER_mu_d_paris],t,calc->acc[ORDER_mu_d_paris])') self.assertEqual(jac['caches'][jac['jac'][2][0]], '-X[ORDER_S_nyc]*diffed[ORDER_diff__r0_nyc]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_nyc],t,calc->acc[ORDER_N_nyc])') self.assertEqual(jac['caches'][jac['jac'][3][1]], '-X[ORDER_S_paris]*diffed[ORDER_diff__r0_paris]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_paris],t,calc->acc[ORDER_N_paris])') # testing jac_obs # all_inc self.assertEqual(jac['caches'][jac['jac_obs'][0][0]], 'gsl_spline_eval(calc->spline[ORDER_mu_d_nyc],t,calc->acc[ORDER_mu_d_nyc])+(gsl_vector_get(par,ORDER_v))') self.assertEqual(jac['caches'][jac['jac_obs'][0][1]], 'gsl_spline_eval(calc->spline[ORDER_mu_d_paris],t,calc->acc[ORDER_mu_d_paris])+(gsl_vector_get(par,ORDER_v))') self.assertEqual(jac['caches'][jac['jac_obs_diff'][0][0]['value']], '0') self.assertEqual(jac['caches'][jac['jac_obs_diff'][0][1]['value']], '0') # nyc_inc self.assertEqual(jac['caches'][jac['jac_obs'][1][0]], 'X[ORDER_S_nyc]*diffed[ORDER_diff__r0_nyc]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_nyc],t,calc->acc[ORDER_N_nyc])') self.assertEqual(jac['caches'][jac['jac_obs'][1][1]], '0') self.assertEqual(jac['caches'][jac['jac_obs_diff'][1][0]['value']], 'X[ORDER_I_nyc]*X[ORDER_S_nyc]*gsl_vector_get(par,ORDER_v)/gsl_spline_eval(calc->spline[ORDER_N_nyc],t,calc->acc[ORDER_N_nyc])') self.assertEqual(jac['caches'][jac['jac_obs_diff'][1][1]['value']], '0') def test_cache_special_function_C(self): caches = map(lambda x: self.m_diff.make_C_term(x, False), ['sin(2*PI*(t +r0))', 'sin(2*PI*(t +r0))', 'sin(2*PI*(t +r0)) + correct_rate(v)']) sf = self.m_diff.cache_special_function_C(caches) self.assertEqual(sf, ['sin(2*PI*(r0+t))', 'ssm_correct_rate(gsl_vector_get(par,ORDER_v),dt)']) self.assertEqual(caches, ['_sf[0]', '_sf[0]', '_sf[1]+_sf[0]']) def test_cache_special_function_C_nested(self): caches = map(lambda x: self.m_diff.make_C_term(x, False), ['pow(correct_rate(correct_rate(v)), pow(2,4))']) sf = self.m_diff.cache_special_function_C(caches) self.assertEqual(sf, ['pow(ssm_correct_rate(ssm_correct_rate(gsl_vector_get(par,ORDER_v),dt),dt),pow(2,4))']) self.assertEqual(caches, ['_sf[0]']) if __name__ == '__main__': unittest.main()