nandetrend
Module for detrending data arrays containing NaN values.
This module provides functionality to remove linear trends from data arrays while properly handling NaN values. It supports:
Automatic detection of NaN values
Linear trend removal using robust fitting
Fallback to scipy.signal.detrend for NaN-free data
Functions
Remove linear trend from array while handling NaN values. |
- nandetrend.nandetrend(x)[source]
Remove linear trend from array while handling NaN values.
This function removes linear trends from data arrays that may contain NaN values. It uses a robust fitting method when NaNs are present, and falls back to scipy.signal.detrend for NaN-free data.
- Parameters:
x (
ndarray
) – 1-dimensional numpy array to detrend. May contain NaN values.- Returns:
Array of same shape as input with linear trend removed. NaN values in input remain NaN in output.
- Return type:
ndarray
Notes
The detrending process: 1. Check for presence of NaN values 2. If NaNs present:
Fit linear trend using nanlinfit
Subtract trend from data
If no NaNs: - Use scipy.signal.detrend