environmentaltools.spatial.interp

environmentaltools.spatial.interp(base, data, dist=100, method='linear', fill_values=nan)[source]

Interpolate spatial data onto a regular grid or specified points.

Performs 2D spatial interpolation from scattered base data points to either a regular grid or specified target points. Uses scipy.interpolate.griddata for the interpolation.

Parameters:

  • base (pd.DataFrame) – Base spatial data for interpolation. Must contain columns ‘x’, ‘y’, and ‘z’ where z is the variable to interpolate.

  • data (pd.DataFrame or dict) – Target interpolation domain. If DataFrame with two rows, defines rectangular bounds [xmin, xmax], [ymin, ymax] for regular grid creation. If dict with ‘x’ and ‘y’ keys, contains target point coordinates for interpolation.

  • dist (float, optional) – Grid spacing in same units as coordinates when creating regular grid. Only used if data is DataFrame with bounds. Default is 100.

  • method (str, optional) – Interpolation method: ‘linear’, ‘nearest’, or ‘cubic’. Default is ‘linear’. See scipy.interpolate.griddata for details.

  • fill_values (float, optional) – Value used to fill points outside the convex hull of base data. Default is np.nan.

Returns:

  • x (np.ndarray) – X coordinates of interpolation points (2D meshgrid or flattened array)

  • y (np.ndarray) – Y coordinates of interpolation points (2D meshgrid or flattened array)

  • z (np.ndarray) – Interpolated values at (x, y) locations

  • df (pd.DataFrame) – Interpolation results as DataFrame with columns [‘x’, ‘y’, ‘z’]

Examples

>>> import pandas as pd
>>> import numpy as np
>>> # Create base data
>>> base = pd.DataFrame({'x': [0, 1, 2], 'y': [0, 1, 2], 'z': [10, 20, 30]})
>>> # Define interpolation bounds
>>> bounds = pd.DataFrame({'x': [0, 2], 'y': [0, 2]})
>>> x, y, z, df = interp(base, bounds, dist=0.5)