Normal Variable

class uqpce.pce.variables.continuous.NormalVariable(mean, stdev, order=2, name='', number=0)

Represents a normal variable. The methods in this class correspond to those of a normal variable.

Parameters:

• mean – the mean of the variable

• stdev – the standard deviation of the variable

• order – the order of the model to calculate the orthogonal polynomials and norm squared values

• name – the name of the variable

• number – the number of the variable from the file

check_distribution(X)

Overrides the Variable class check_distribution to align with a normal distribution.

Parameters:

X – The array of samples to check against the variable distribution

check_num_string()

Searches to replace sring ‘pi’ with its numpy equivalent in any of the values that might contain it.

generate_orthopoly()

Generates the orthogonal polynomials for a normal variable up to the order of polynomial expansion.

generate_samples(count, standardize=False)

Overrides the Variable class generate_samples to align with a normal distribution.

Parameters:

• count – the number of points needed to be generated

• standardize – boolean for generating samples for the standardized version of the variable

get_mean()

Return the mean of the variable.

get_norm_sq_val(matrix_value)

Overrides the Variable class get_norm_sq_val to align with a normal distribution.

Parameters:

matrix_val – the value in the model matrix to consider

standardize(orig, std_vals)

Overrides the Variable class standardize to align with a normal distribution.

Parameters:

• orig – the un-standardized values

• std_vals – the attribue name for the standardized vals

standardize_points(values)

Standardizes and returns the inputs points.

Parameters:

values – unstandardized points corresponding to the variable’s distribution

unstandardize_points(value)

Calculates and returns the unscaled variable value from the standardized value.

Parameters:

value – the standardized value to be unstandardized