Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
import numpy as np | |
from scipy.signal import argrelextrema | |
def find_extrema(x, y): | |
# Find the indices of the local maxima and minima | |
maxima_idx = argrelextrema(x, np.greater)[0] | |
minima_idx = argrelextrema(x, np.less)[0] | |
# Get the x and y values at the extrema | |
maxima_x = x[maxima_idx] | |
maxima_y = y[maxima_idx] | |
minima_x = x[minima_idx] | |
minima_y = y[minima_idx] | |
# Combine the maxima and minima into a single array | |
extrema_x = np.concatenate((maxima_x, minima_x)) | |
extrema_y = np.concatenate((maxima_y, minima_y)) | |
# Determine whether each extrema is a maximum or minimum | |
is_maxima = np.zeros(len(extrema_x), dtype=bool) | |
is_maxima[:len(maxima_x)] = True | |
# Sort the extrema by x-value | |
idx = np.argsort(extrema_x) | |
extrema_x = extrema_x[idx] | |
extrema_y = extrema_y[idx] | |
is_maxima = is_maxima[idx] | |
# Convert the boolean array to a list of strings | |
extrema_type = ["arrow-left" if is_max else "arrow-right" for is_max in is_maxima] | |
return extrema_x, extrema_y, extrema_type | |