class ModelTest():
def __init__(self, cls, name):
'''
cls: model class
name: string to pass to FitWrapper
'''
self.cls = cls()
self.name = name
def generate_data(self, xmin, xmax, steps, noise, true_params):
'''
Generate noisy data around the model
'''
class dataset():
def __init__(self, x, y):
N = len(x)
self.data = np.zeros((N, 2))
self.data[:, 0] = x
self.data[:, 1] = y
x = np.linspace(xmin, xmax, steps)
self.true_params = true_params
y = self.cls.model(x,
true_params) + noise * np.random.normal(size=len(x))
ds = dataset(x, y)
self.fw = FitWrapper(ds, 0)
self.fw.setModel(self.name)
for p in self.fw.getParameters():
self.fw.getManualValue(p) # force guess of initial parameters
self.x = x
self.y = y
def fit(self):
self.fw.doFit()
def plot(self, fit=True):
plt.plot(self.x, self.y, 'ro')
if fit:
output = self.fw.evaluateFittedParameters()
plt.plot(output[:, 0], output[:, 1])
plt.show()
def print_results(self):
print '***** FIT RESULTS *****'
print '{}\t{}\t{}'.format('PARAM', 'TRUE', 'FITTED')
for i, p in enumerate(self.fw.getParameters()):
print '{}\t{}\t{}'.format(p, self.true_params[i],
self.fw.getFittedValue(p))
class ModelTest():
def __init__(self, cls, name):
'''
cls: model class
name: string to pass to FitWrapper
'''
self.cls = cls()
self.name = name
def generate_data(self, xmin, xmax, steps, noise, true_params):
'''
Generate noisy data around the model
'''
class dataset():
def __init__(self, x, y):
N = len(x)
self.data = np.zeros((N,2))
self.data[:,0] = x
self.data[:,1] = y
x = np.linspace(xmin, xmax, steps)
self.true_params = true_params
y = self.cls.model(x, true_params) + noise*np.random.normal(size=len(x))
ds = dataset(x, y)
self.fw = FitWrapper(ds, 0)
self.fw.setModel(self.name)
for p in self.fw.getParameters():
self.fw.getManualValue(p) # force guess of initial parameters
self.x = x
self.y = y
def fit(self):
self.fw.doFit()
def plot(self, fit = True):
plt.plot(self.x, self.y, 'ro')
if fit:
output = self.fw.evaluateFittedParameters()
plt.plot(output[:,0], output[:,1])
plt.show()
def print_results(self):
print '***** FIT RESULTS *****'
print '{}\t{}\t{}'.format('PARAM','TRUE','FITTED')
for i, p in enumerate(self.fw.getParameters()):
print '{}\t{}\t{}'.format(p, self.true_params[i], self.fw.getFittedValue(p))
def generate_data(self, xmin, xmax, steps, noise, true_params):
'''
Generate noisy data around the model
'''
class dataset():
def __init__(self, x, y):
N = len(x)
self.data = np.zeros((N, 2))
self.data[:, 0] = x
self.data[:, 1] = y
x = np.linspace(xmin, xmax, steps)
self.true_params = true_params
y = self.cls.model(x,
true_params) + noise * np.random.normal(size=len(x))
ds = dataset(x, y)
self.fw = FitWrapper(ds, 0)
self.fw.setModel(self.name)
for p in self.fw.getParameters():
self.fw.getManualValue(p) # force guess of initial parameters
self.x = x
self.y = y
def generate_data(self, xmin, xmax, steps, noise, true_params):
'''
Generate noisy data around the model
'''
class dataset():
def __init__(self, x, y):
N = len(x)
self.data = np.zeros((N,2))
self.data[:,0] = x
self.data[:,1] = y
x = np.linspace(xmin, xmax, steps)
self.true_params = true_params
y = self.cls.model(x, true_params) + noise*np.random.normal(size=len(x))
ds = dataset(x, y)
self.fw = FitWrapper(ds, 0)
self.fw.setModel(self.name)
for p in self.fw.getParameters():
self.fw.getManualValue(p) # force guess of initial parameters
self.x = x
self.y = y
