pyzebra/pyzebra/ccl_findpeaks.py

76 lines
3.3 KiB
Python

import numpy as np
import scipy as sc
from scipy.interpolate import interp1d
from scipy.signal import savgol_filter
def ccl_findpeaks(
scan, int_threshold=0.8, prominence=50, smooth=False, window_size=7, poly_order=3
):
"""function iterates through the dictionary created by load_cclv2 and locates peaks for each scan
args: scan - a single scan,
int_threshold - fraction of threshold_intensity/max_intensity, must be positive num between 0 and 1
i.e. will only detect peaks above 75% of max intensity
prominence - defines a drop of values that must be between two peaks, must be positive number
i.e. if promimence is 20, it will detect two neigbouring peaks of 300 and 310 intesities,
if none of the itermediate values are lower that 290
smooth - if true, smooths data by savitzky golay filter, if false - no smoothing
window_size - window size for savgol filter, must be odd positive integer
poly_order = order of the polynomial used in savgol filter, must be positive integer smaller than
window_size returns: dictionary with following structure:
D{M34{ 'num_of_peaks': 1, #num of peaks
'peak_indexes': [20], # index of peaks in omega array
'peak_heights': [90.], # height of the peaks (if data vere smoothed
its the heigh of the peaks in smoothed data)
"""
if not 0 <= int_threshold <= 1:
int_threshold = 0.8
print(
"Invalid value for int_threshold, select value between 0 and 1, new value set to:",
int_threshold,
)
if not isinstance(window_size, int) or (window_size % 2) == 0 or window_size <= 1:
window_size = 7
print(
"Invalid value for window_size, select positive odd integer, new value set to!:",
window_size,
)
if not isinstance(poly_order, int) or window_size < poly_order:
poly_order = 3
print(
"Invalid value for poly_order, select positive integer smaller than window_size, new value set to:",
poly_order,
)
if not isinstance(prominence, (int, float)) and prominence < 0:
prominence = 50
print("Invalid value for prominence, select positive number, new value set to:", prominence)
omega = scan["om"]
counts = np.array(scan["Counts"])
if smooth:
itp = interp1d(omega, counts, kind="linear")
absintensity = [abs(number) for number in counts]
lowest_intensity = min(absintensity)
counts[counts < 0] = lowest_intensity
smooth_peaks = savgol_filter(itp(omega), window_size, poly_order)
else:
smooth_peaks = counts
peaks, properties = sc.signal.find_peaks(
smooth_peaks, height=int_threshold * max(smooth_peaks), prominence=prominence
)
scan["num_of_peaks"] = len(peaks)
scan["peak_indexes"] = peaks
scan["peak_heights"] = properties["peak_heights"]
scan["smooth_peaks"] = smooth_peaks # smoothed curve