polire

A collection of interpolation methods.


Keywords
python, interpolation, gis, spatial, spatial-analysis, sensor-placement, placement
License
BSD-3-Clause
Install
pip install polire==0.1.3

Documentation

example workflow

Polire

The word "interpolation" has Latin origin and is composed of two words - Inter meaning between and Polire meaning to polish.

This repository is a collection of several spatial interpolation algorithms.

Examples

Minimal example of interpolation

import numpy as np
from polire import Kriging

# Data
X = np.random.rand(10, 2) # Spatial 2D points
y = np.random.rand(10) # Observations
X_new = np.random.rand(100, 2) # New spatial points

# Fit
model = Kriging()
model.fit(X, y)

# Predict
y_new = model.predict(X_new)

Supported Interpolation Methods

from polire import (
    Kriging, # Best spatial unbiased predictor
    GP, # Gaussian process interpolator from GPy
    IDW, # Inverse distance weighting
    SpatialAverage,
    Spline,
    Trend,
    Random, # Predict uniformly within the observation range, a reasonable baseline
    NaturalNeighbor,
    CustomInterpolator # Supports any regressor from Scikit-learn
)

Use GP kernels from GPy and regressors from sklearn

from sklearn.linear_model import LinearRegression # or any Scikit-learn regressor
from GPy.kern import Matern32 # or any other GPy kernel

from polire import GP, CustomInterpolator

# GP model
model = GP(Matern32(input_dim=2))

# Sklearn model
model = CustomInterpolator(LinearRegression(normalize = True))

More info

Contributors: S Deepak Narayanan, Zeel B Patel, Apoorv Agnihotri, and Nipun Batra.

This project is a part of Sustainability Lab at IIT Gandhinagar.

Acknowledgement to sklearn template for helping to package into a PiPy package.