def get_symbolic_model(f, dim_x, n_points=100, x_range=[0, 1]):
H = load_H()
loss_tol = load_hyperparameters()['loss_tol']
faithful_models = []
losses = []
for k in range(len(H)):
for l in range(len(H)):
print("===================================================================================================")
print("Testing Hyperparameter Configuration k = ", k + 1 , " ; l = ", l+1)
faithful_model, loss = symbolic_modeling(f, dim_x,
H['hyper_' + str(k + 1)][1], H['hyper_' + str(l + 1)][1],
H['hyper_' + str(k + 1)][0], H['hyper_' + str(l + 1)][0],
n_points=n_points, x_range=x_range)
faithful_models.append(faithful_model)
losses.append(loss)
if losses[-1] <= loss_tol:
print("==========The desired loss was achieved so the algorithm stopped==========")
break
best_model = np.argmin(np.array(losses))
X = np.random.uniform(x_range[0], x_range[1], dim_x * n_points)
X = X.reshape((n_points, dim_x))
Y_true = f(X).reshape((-1, 1))
Y_est = faithful_models[best_model].evaluate(X).reshape((-1, 1))
R2_perf = compute_Rsquared(Y_true, Y_est)
return faithful_models[best_model], R2_perf
def symbolic_regressor(f, npoints, xrange):
X = np.linspace(xrange[0], xrange[1], npoints).reshape((-1, 1))
y = f(X)
est_gp = SymbolicRegressor(population_size=5000,
generations=20, stopping_criteria=0.01,
p_crossover=0.7, p_subtree_mutation=0.1,
p_hoist_mutation=0.05, p_point_mutation=0.1,
max_samples=0.9, verbose=1,
parsimony_coefficient=0.01, random_state=0)
est_gp.fit(X, y)
sym_expr = str(est_gp._program)
converter = {
'sub': lambda x, y: x - y,
'div': lambda x, y: x / y,
'mul': lambda x, y: x * y,
'add': lambda x, y: x + y,
'neg': lambda x: -x,
'pow': lambda x, y: x ** y
}
x, X0 = symbols('x X0')
sym_reg = simplify(sympify(sym_expr, locals=converter))
sym_reg = sym_reg.subs(X0, x)
Y_true = y.reshape((-1, 1))
Y_est = np.array([sympify(str(sym_reg)).subs(x, X[k]) for k in range(len(X))]).reshape((-1, 1))
R2_perf = compute_Rsquared(Y_true, Y_est)
return sym_reg, R2_perf
def get_symbolic_model(f, npoints, xrange):
hyperparameter_space = load_hyperparameter_config()
loss_threshold = 10e-5
symbol_exprs = []
losses_ = []
for k in range(len(hyperparameter_space)):
symbolic_model, Loss_ = symbolic_modeling(
f, hyperparameter_space['hyper_' + str(k + 1)][1],
hyperparameter_space['hyper_' + str(k + 1)][0], npoints, xrange)
symbol_exprs.append(symbolic_model)
losses_.append(Loss_)
if losses_[-1] <= loss_threshold:
break
best_model = np.argmin(np.array(losses_))
X = np.linspace(xrange[0], xrange[1], npoints).reshape((-1, 1))
Y_true = f(X).reshape((-1, 1))
Y_est = symbol_exprs[best_model].evaluate(X).reshape((-1, 1))
R2_perf = compute_Rsquared(Y_true, Y_est)
return symbol_exprs[best_model], R2_perf
