def compute_func_values(cell, theta, model, fobj, f, options, istrain):
'''
Compute the objective function value, correlation, explained variance
Outputs
-------
fun (double):
objective value
cc (double):
correlation coefficient
ev (double):
explained variance
'''
if istrain:
data = ccls.get_data(cell, model, options)
else:
temp_options = {}
for key in options.keys():
if key == 'crossval':
temp_options[key] = False
else:
temp_options[key] = options[key]
data = ccls.get_data(cell, model, temp_options)
options['basis'] = lnks.LinearFilterBasis_8param()
options['stim'] = data[0] - np.mean(data[0])
if options['is_grad']:
fun, grad = fobj(theta, data[0], data[1], options)
else:
fun = fobj(theta, data[0], data[1], options)
if model.lower() == 'lnks_mp':
y = data[1][1]
v, y_est = f(theta, data[0], options)
else:
y = data[1]
y_est = f(theta, data[0], options)
if istrain:
cc = stats.corrcoef(y, y_est)
ev = stats.variance_explained(y, y_est)
else:
cc = stats.corrcoef(y[-20000:], y_est[-20000:])
ev = stats.variance_explained(y[-20000:], y_est[-20000:])
return fun, cc, ev
def optimize(fobj, f, theta, data, bnds=None, options=None):
'''
Optimization using scipy.optimize.minimize module
Input
-----
fobj (function):
objective function
f (function):
model function
theta (ndarray):
initial parameter
data (tuple of ndarray):
input and output data
bnds (tuple of tuples or None):
Boundaries of model parameters
options (dictionary)
grad (Bool):
True if using gradient for optimization, else False
pathway (int):
LNK model pathway (1, 2, otherwise None)
Output
------
result
theta: estimated parameter
success: optimization converged
fun: objective value
corrcoef: correlation coefficient
evar: explained variance
theta_init: initial theta
jac: gradient
'''
theta_init = theta
if bnds:
if options['is_grad']:
res = minimize(fobj, theta_init, args=data, method='L-BFGS-B', jac=True, bounds=bnds, options={'disp':DISP, 'maxiter':MAX_ITER})
else:
res = minimize(fobj, theta_init, args=data, method='L-BFGS-B',bounds=bnds, options={'disp':DISP, 'maxiter':MAX_ITER})
else:
# optimization
if options['is_grad']:
res = minimize(fobj, theta_init, args=data, method='L-BFGS-B', jac=True, options={'disp':DISP, 'maxiter':MAX_ITER})
else:
res = minimize(fobj, theta_init, args=data, method='L-BFGS-B', options={'disp':DISP, 'maxiter':MAX_ITER})
model = options['model']
if model.lower() == 'lnks_mp':
y = data[1][1]
# v, y_est = f(theta, data[0], options['pathway'])
v, y_est = f(theta, data[0], options) # fast_lnks_objective
else:
y = data[1]
# y_est = f(theta, data[0], options['pathway'])
y_est = f(theta, data[0], options) # fast_lnks_objective
cc = _stats.corrcoef(y_est, y)
ev = _stats.variance_explained(y_est, y)
# results
if options['is_grad']:
jac = res.jac
else:
jac = None
result = {
"theta": res.x,
"success": res.success,
"fun": res.fun,
"corrcoef": cc,
"evar": ev,
"theta_init": theta_init,
"jac": jac}
return result
