Mann-Kendall Trend Test
Source: src/mann_kendall.rs
This module provides an implementation of the Mann-Kendall trend test for Polars Series.
The Mann-Kendall test checks for a trend in a time series by comparing each data point
to subsequent data points. It computes a statistic S indicating how many pairs are
in increasing vs. decreasing order and then normalizes it via
[ S / (0.5 * n * (n - 1)) ], matching standard Mann-Kendall definitions.
Python API
mann_kendall
Apply the Mann-Kendall trend test to the first Series in inputs.
Parameters
| Parameter | Description |
|---|---|
inputs |
a slice of Series, with the first entry expected to be a float Series (f64). |
Returns
A single-valued Float64 Series containing the normalized Mann-Kendall statistic.
This matches the Python version mk_stat = S / (0.5 * n * (n - 1)).
Errors
Returns an error if the first Series is not of f64 type or if it contains nulls.
Internal Functions
These are internal Rust functions not directly callable from Python.
query
Query how many increments occur up to and including index i in the Fenwick tree.