Taweret.models namespace

Submodules

class Taweret.models.coleman_models.coleman_model_1[source]

Bases: BaseModel

evaluate(input_values: array, model_param: array, full_corr=False) → array[source]

Predict the mean and error for given input values

Parameters:

input_values (numpy 1darray) -- input parameter values
model_param (numpy 1darray) -- value of the model parameter

log_likelihood_elementwise(x_exp, y_exp, y_err, model_param)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

property prior

set_prior(bilby_priors=None)[source]: Set the prior on model parameters.

class Taweret.models.coleman_models.coleman_model_2[source]

Bases: BaseModel

evaluate(input_values: array, model_param: array, full_corr=False) → array[source]

Predict the mean and error for given input values

Parameters:

input_values (numpy 1darray) -- input parameter values
model_param (numpy 1darray) -- value of the model parameter

log_likelihood_elementwise(x_exp, y_exp, y_err, model_param)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

property prior

set_prior(bilby_priors=None)[source]: Set the prior on model parameters.

class Taweret.models.coleman_models.coleman_truth[source]

Bases: BaseModel

evaluate(input_values: array) → array[source]

Predict the mean and error for given input values

Parameters:: input_values (numpy 1darray) -- input parameter values

log_likelihood_elementwise(x_exp, y_exp, y_err, model_param)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

set_prior(bilby_priors=None)[source]

User must provide function that sets a member variable called _prior. Dictionary of prior distributions. Format should be compatible with sampler.

Example:

class MyModel(BaseMixer):
    # . . .
    def set_prior(self, prior_dict):
        self._prior = prior_dict
    # . . .

class Taweret.models.polynomial_models.cos_exp(k, x0)[source]

Bases: BaseModel

Cosine Taylor series expansion model class.

Parameters:

param int k:: the degree of the expansion.
param float x0:: the center of the expansion.

evaluate(x)[source]

Evaluate the Taylor series at a grid of x's. The standard deviation output is set to 1 by default.

Parameters:

param np.ndarray x:: design matrix.

Returns:

returns:: mean and standard deviation of the model at the x grid points.
rtype:: np.ndarray, np.ndarray
return values:: mean predictions.
return values:: standard deviation of the predictions.

log_likelihood_elementwise()[source]: Obtain the log likelihood for the model. Not needed for this model.

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.

class Taweret.models.polynomial_models.polynomal_model(a=0, b=0, c=1, p=1)[source]

Bases: BaseModel

Polynomial models class. Used to define a function of the form

\[f(x) = c(x-a)^p + b\]

Parameters:

param float a:: center parameter.
param float b:: shift parameter.
param float c:: scale parameter.
param float p:: power parameter.

evaluate(x)[source]

Evaluate the polynomial at a grid of x's. The standard deviation output is set to 1 by default.

Parameters:

param np.ndarray x:: design matrix.

Returns:

returns:: mean and standard deviation of the model at the x grid points.
rtype:: np.ndarray, np.ndarray
return values:: mean predictions.
return values:: standard deviation of the predictions.

log_likelihood_elementwise()[source]: Obtain the log likelihood for the model. Not needed for this model.

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.

class Taweret.models.polynomial_models.sin_cos_exp(ks, kc, xs, xc)[source]

Bases: BaseModel

Taylor series expansion of

\[f(x) = \sin(x_1) + \cos(x_2)\]

Parameters:

param int ks:: the degree of the sine expansion.
param int kc:: the degree of the cosine expansion.
param float xs:: the center of the sine expansion.
param float xc:: the center of the cosine expansion.

evaluate(x)[source]

Evaluate the model at a grid of x's. The standard deviation output is set to 1 by default.

Parameters:

param np.ndarray x:: design matrix.

Returns:

returns:: mean and standard deviation of the model at the x grid points.
rtype:: np.ndarray, np.ndarray
return values:: mean predictions.
return values:: standard deviation of the predictions.

log_likelihood_elementwise()[source]: Obtain the log likelihood for the model. Not needed for this model.

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.

class Taweret.models.polynomial_models.sin_exp(k, x0)[source]

Bases: BaseModel

Sine Taylor series expansion model class.

Parameters:

param int k:: the degree of the expansion.
param float x0:: the center of the expansion.

evaluate(x)[source]

Evaluate the Taylor Series at a grid of x's. The standard deviation output is set to 1 by default.

Parameters:

param np.ndarray x:: design matrix.

Returns:

returns:: mean and standard deviation of the model at the x grid points.
rtype:: np.ndarray, np.ndarray
return values:: mean predictions.
return values:: standard deviation of the predictions.

log_likelihood_elementwise()[source]: Obtain the log likelihood for the model. Not needed for this model.

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.

class Taweret.models.samba_models.Data[source]

Bases: BaseModel

evaluate(input_values: array, error=0.01) → array[source]

Evaluate the data and error for given input values

Parameters:

input_valuesnumpy 1darray: coupling strength (g) values for data generation
errorfloat: defines the relative error as a fraction between (0,1)

Returns:

datanumpy 1darray: The array of data points
sigmanumpy 1darray: The errors on each data point

log_likelihood_elementwise()[source]: Obtain the log likelihood for the model. Not needed for this model.

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.

class Taweret.models.samba_models.Highorder(order, error_model='informative')[source]

Bases: BaseModel

The SAMBA highorder series expansion function.

Parameters:

orderint: Truncation order of expansion
error_modelstr: Error calculation method. Either 'informative' or 'uninformative'

raises TypeError:: If the order is not an integer

evaluate(input_values: array) → array[source]

Evaluate the mean and standard deviation for given input values

Parameters:

input_valuesnumpy 1darray: coupling strength (g) values

Returns:

meannumpy 1darray: The mean of the model
np.sqrt(var)numpy 1darray: The truncation error of the model

log_likelihood_elementwise(x_exp, y_exp, y_err, model_param)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

set_prior()[source]: Set the prior on the model parameters. Not needed for this model.

class Taweret.models.samba_models.Loworder(order, error_model='informative')[source]

Bases: BaseModel

The SAMBA loworder series expansion function. This model has been previously calibrated.

Parameters:

orderint: Truncation order of expansion
error_modelstr: Error calculation method. Either 'informative' or 'uninformative'

raises TypeError:: If the order is not an integer

evaluate(input_values: array) → array[source]

Evaluate the mean and standard deviation for given input values to the function

Parameters:

input_valuesnumpy 1darray: coupling strength (g) values

Returns:

meannumpy 1darray: The mean of the model
np.sqrt(var)numpy 1darray: The truncation error of the model

log_likelihood_elementwise(x_exp, y_exp, y_err, model_param)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

set_prior()[source]: Set the prior on model parameters. Not needed for this model.

class Taweret.models.samba_models.TrueModel[source]

Bases: BaseModel

evaluate(input_values: array) → array[source]

Evaluate the mean of the true model for given input values.

Parameters:

input_valuesnumpy 1darray: coupling strength (g) values

Returns:

meannumpy 1darray: The true model evaluated at each point of the given input space
np.sqrt(var)numpy 1darray: The standard deviation of the true model. This will obviously be an array of zeros.

log_likelihood_elementwise(x_exp, y_exp, y_err)[source]

Calculate log_likelihood for array of points given

Returns:

log_likelisnp.ndarray: an array of length as shape[0] of the input evaluation points

Example:

class MyModel(BaseModel):
    def log_likelihood_elementwise(
        self, y_exp, y_err, model_params
    ):
        # Assuming a normal distribution for error
        y = self.evaluate(model_params)
        # If y_exp, y_err, y are numpy arrays of same length
        return np.exp(-(y - y_exp) **2 / (2 * y_err ** 2)) \
            / np.sqrt(2 * np.pi * y_err ** 2))

set_prior()[source]: Set the prior on any model parameters. Not needed for this model.