SciFit

Comprehensive fits for scientists

Welcome to SciFit project the Python package for comprehensive fits for scientists designed to ease fitting procedure and automatically perform the quality assessment.

The SciFit project aims to support your work by:

• Providing a clean, stable and compliant interface for each solver;
• Perform ad hoc transformations, processing and tests on each stage of a solver procedure;
• Render high quality figures summarizing solver solution and the quality assessment.

Installation

You can install the SciFit package by issuing:

python -m pip install --upgrade scifit

Which update you to the latest version of the package.

Quick start

Let's fit some data:

from scifit.solvers.scientific import *

# Select a specific solver:
solver = HillEquationFitSolver()

# Create some synthetic dataset:
data = solver.synthetic_dataset(
    xmin=0.0, xmax=5.0, resolution=50,
    parameters=[3.75, 0.21],
    sigma=0.1, scale_mode="auto", seed=1234,
)

# Perform regression:
solution = solver.fit(data)

# Render results:
axe = solver.plot_fit()

We have a nice adjustments for such noisy data:

Or even better a full fit report to check each key points at once:

solver.report("hill_report")

Which produces a PDF file called hill_report.pdf.

Resources

• Documentations
• Repository