Continuous Fit Implementation

A continuous fit can be performed by instantiating the Phitter class from the Phitter library. This class provides several parameters, each of which is summarized below:

data (list[int | float] or numpy.ndarray):
The dataset on which distributions are to be fitted.
fit_type (str, optional; default: "continuous"):
Indicates the type of distribution fitting. For continuous datasets, this parameter should be set to "continuous".
num_bins (int, optional):
Specifies the number of bins to be used for histogram creation. If not defined, Phitter automatically determines the optimal bin count by applying Doane’s rule.
confidence_level (float, optional; default: 0.95):
Denotes the confidence level for statistical tests designed to assess goodness-of-fit.
minimum_sse (float, optional; default: numpy.inf):
Establishes a threshold for the minimum sum of squared errors (SSE). Any distribution whose SSE surpasses this value is excluded from consideration.
subsample_size (int, optional; default: None):
Specifies the size of a randomly drawn subset of the dataset, if subsampling is necessary for distribution fitting.
subsample_estimation_size (int, optional; default: None):
Determines the size of a random subset used exclusively to estimate parameters for each distribution, if parameter estimation subsampling is required.
distributions_to_fit (list[str] or "all", optional; default: "all"):
Identifies which distributions are evaluated during fitting. Accepts either a list of specific distributions (see distributions documentation) or the string "all" to fit every available distribution.
exclude_distributions (list[str] or str, optional; default: "any"):
Specifies distributions that must be explicitly excluded from fitting. If set to "any", no distributions are excluded.

Example of a basic continuous fit
When no additional parameters are specified, the following code performs a simple continuous fit:

python

import phitter

# Define the dataset
data: list[int | float] = [...]

# Perform a continuous fit using default settings
phi = phitter.Phitter(data=data)
phi.fit(n_workers=4)

Example of a more comprehensive continuous fit
The code snippet below demonstrates how to configure additional parameters, such as the confidence level or minimum SSE threshold:

python

import phitter

# Define the dataset
data: list[int | float] = [...]

# Execute a continuous fit with specified parameters
phi = phitter.Phitter(
    data=data,
    fit_type="continuous",
    num_bins=15,
    confidence_level=0.95,
    minimum_sse=1e-2,
    distributions_to_fit=["beta", "normal", "fatigue_life", "triangular"]
)
phi.fit(n_workers=4)

After creating and initializing the Phitter class with the chosen parameters, the fitting procedure is started by invoking the .fit() method. This method systematically evaluates each specified distribution, computes various goodness-of-fit measures, and then ranks the fitted distributions according to the sum of squared errors and test rejections.

Continuous Fit Implementation ​

Continuous Fit Implementation