[DOC,MNT,BUG] Documentation and testing for downsample.py (#45)

* 1.Fixes a bug in downsample.py when
n_timepoints / new_ratio is an integer
2.complete a testing for downsample.py
in transformations/tests/test_downsample.py
3.write documentation for transformations/downsample.py

* 1.Fixes a bug in downsample.py
when n_timepoints / new_ratio is an integer
2.complete a testing for downsample.py
in transformations/tests/test_downsample.py
3.write documentation for transformations/downsample.py

* modify the variable names
both in the code and comments

---------

Co-authored-by: futuer <[email protected]>

aeon_neuro/transformations/downsample.py

@@ -1,13 +1,32 @@
+"""Downsample series by frequency."""
+
 import numpy as np
 
 
 def downsample_series(X, sfreq, target_sample_rate):
+ """Downsample a time series.
+
+ Parameters
+ ----------
+ X : a numpy array of shape = [n_cases, n_channels, n_timepoints]
+ The input time series data to be downsampled.
+ sfreq : int
+ The original sampling frequency of the time series data.
+ target_sample_rate : int
+ The target sampling frequency after downsampling.
+
+ Returns
+ -------
+ downsampled : a numpy array of
+ shape = [n_cases, n_channels, updated_timepoints]
+ The downsampled time series data.
+ """
  new_ratio = int(sfreq / target_sample_rate)
- n_instances, n_dimensions, n_timepoints = np.shape(X)
- updated_timepoints = int(n_timepoints / new_ratio)
- downsampled_data = np.zeros((n_instances, n_dimensions, updated_timepoints + 1))
- for i in range(n_instances):
- for j in range(n_dimensions):
+ n_cases, n_channels, n_timepoints = np.shape(X)
+ updated_timepoints = int(np.ceil(n_timepoints / new_ratio))
+ downsampled_data = np.zeros((n_cases, n_channels, updated_timepoints))
+ for i in range(n_cases):
+ for j in range(n_channels):
  updated_index = 0
  for k in range(0, n_timepoints, new_ratio):
  downsampled_data[i][j][updated_index] = X[i][j][k]

aeon_neuro/transformations/tests/__init__.py

@@ -0,0 +1 @@
+"""Unit tests of transformations."""

aeon_neuro/transformations/tests/test_downsample.py

@@ -0,0 +1,46 @@
+"""Tests for downsample_series."""
+
+import numpy as np
+import pytest
+
+from aeon_neuro.transformations.downsample import downsample_series
+
+X = np.array(
+ [
+ [[1, 2, 3, 4, 5, 6, 7, 8, 9, 10], [10, 9, 8, 7, 6, 5, 4, 3, 2, 1]],
+ [
+ [10, 20, 30, 40, 50, 60, 70, 80, 90, 100],
+ [100, 90, 80, 70, 60, 50, 40, 30, 20, 10],
+ ],
+ ]
+)
+
+testdata = [
+ (
+ X,
+ 10,
+ 5,
+ np.array(
+ [
+ [[1, 3, 5, 7, 9], [10, 8, 6, 4, 2]],
+ [[10, 30, 50, 70, 90], [100, 80, 60, 40, 20]],
+ ]
+ ),
+ ),
+ (
+ X,
+ 10,
+ 3,
+ np.array(
+ [[[1, 4, 7, 10], [10, 7, 4, 1]], [[10, 40, 70, 100], [100, 70, 40, 10]]]
+ ),
+ ),
+ (X, 10, 2, np.array([[[1, 6], [10, 5]], [[10, 60], [100, 50]]])),
+]
+
+
+@pytest.mark.parametrize("X, sfreq, target_sample_rate, expected", testdata)
+def test_downsample_series(X, sfreq, target_sample_rate, expected):
+ """Test the downsampling of a time series."""
+ result = downsample_series(X, sfreq, target_sample_rate)
+ np.testing.assert_array_equal(result, expected)