import itertools import sys import os import argparse import torch from torch import nn import torch.nn.functional as F import torchvision.transforms as T import numpy as np import matplotlib.pyplot as plt import matplotlib.gridspec as gridspec import pandas as pd from pathlib import Path from tqdm.auto import tqdm def load_spec_data(folder: str): """ Load spectrogram of a given electrode-grid recording generated with the wavetracker package. The spectrograms may be to large to load in total, thats why memmory mapping is used (numpy.memmap). Parameters ---------- folder: str Folder where fine spec numpy files generated for grid recordings with the wavetracker package can be found. Returns ------- fill_freqs: ndarray Freuqencies corresponding to 1st dimension of the spectrogram. fill_times: ndarray Times corresponding to the 2nd dimenstion if the spectrigram. fill_spec: ndarray Spectrigram of the recording refered to in the input folder. """ fill_freqs, fill_times, fill_spec = [], [], [] if os.path.exists(os.path.join(folder, 'fill_spec.npy')): fill_freqs = np.load(os.path.join(folder, 'fill_freqs.npy')) fill_times = np.load(os.path.join(folder, 'fill_times.npy')) fill_spec_shape = np.load(os.path.join(folder, 'fill_spec_shape.npy')) fill_spec = np.memmap(os.path.join(folder, 'fill_spec.npy'), dtype='float', mode='r', shape=(fill_spec_shape[0], fill_spec_shape[1]), order='F') elif os.path.exists(os.path.join(folder, 'fine_spec.npy')): fill_freqs = np.load(os.path.join(folder, 'fine_freqs.npy')) fill_times = np.load(os.path.join(folder, 'fine_times.npy')) fill_spec_shape = np.load(os.path.join(folder, 'fine_spec_shape.npy')) fill_spec = np.memmap(os.path.join(folder, 'fine_spec.npy'), dtype='float', mode='r', shape=(fill_spec_shape[0], fill_spec_shape[1]), order='F') return fill_freqs, fill_times, fill_spec def load_tracking_data(folder): base_path = Path(folder) EODf_v = np.load(base_path / 'fund_v.npy') ident_v = np.load(base_path / 'ident_v.npy') idx_v = np.load(base_path / 'idx_v.npy') times_v = np.load(base_path / 'times.npy') return EODf_v, ident_v, idx_v, times_v def load_trial_data(folder): base_path = Path(folder) fish_freq = np.load(base_path / 'analysis' / 'fish_freq.npy') rise_idx = np.load(base_path / 'analysis' / 'rise_idx.npy') rise_size = np.load(base_path / 'analysis' / 'rise_size.npy') fish_baseline_freq = np.load(base_path / 'analysis' / 'baseline_freqs.npy') fish_baseline_freq_time = np.load(base_path / 'analysis' / 'baseline_freq_times.npy') return fish_freq, rise_idx, rise_size, fish_baseline_freq, fish_baseline_freq_time def save_spec_pic(folder, s_trans, times, freq, t_idx0, t_idx1, f_idx0, f_idx1, dataset_folder): size = (7, 7) dpi = 256 f_res, t_res = freq[1] - freq[0], times[1] - times[0] fig_title = (f'{Path(folder).name}__{times[t_idx0]:5.0f}s-{times[t_idx1]:5.0f}s__{freq[f_idx0]:4.0f}-{freq[f_idx1]:4.0f}Hz.png').replace(' ', '0') fig = plt.figure(figsize=(7, 7), num=fig_title) gs = gridspec.GridSpec(1, 1, bottom=0, left=0, right=1, top=1) # ax = fig.add_subplot(gs[0, 0]) ax.imshow(s_trans.squeeze(), cmap='gray', aspect='auto', origin='lower', extent=(times[t_idx0] / 3600, times[t_idx1] / 3600 + t_res, freq[f_idx0], freq[f_idx1] + f_res)) ax.axis(False) plt.savefig(os.path.join(dataset_folder, fig_title), dpi=256) plt.close() return fig_title, (size[0]*dpi, size[1]*dpi) def bboxes_from_file(times_v, fish_freq, rise_idx, rise_size, fish_baseline_freq_time, fish_baseline_freq, pic_save_str, bbox_df, cols, width, height, t0, t1, f0, f1): times_v_idx0, times_v_idx1 = np.argmin(np.abs(times_v - t0)), np.argmin(np.abs(times_v - t1)) for id_idx in range(len(fish_freq)): rise_idx_oi = np.array(rise_idx[id_idx][ (rise_idx[id_idx] >= times_v_idx0) & (rise_idx[id_idx] <= times_v_idx1) & (rise_size[id_idx] >= 10)], dtype=int) rise_size_oi = rise_size[id_idx][(rise_idx[id_idx] >= times_v_idx0) & (rise_idx[id_idx] <= times_v_idx1) & (rise_size[id_idx] >= 10)] if len(rise_idx_oi) == 0: continue closest_baseline_idx = list(map(lambda x: np.argmin(np.abs(fish_baseline_freq_time - x)), times_v[rise_idx_oi])) closest_baseline_freq = fish_baseline_freq[id_idx][closest_baseline_idx] upper_freq_bound = closest_baseline_freq + rise_size_oi lower_freq_bound = closest_baseline_freq left_time_bound = times_v[rise_idx_oi] right_time_bound = np.zeros_like(left_time_bound) for enu, Ct_oi in enumerate(times_v[rise_idx_oi]): Crise_size = rise_size_oi[enu] Cblf = closest_baseline_freq[enu] rise_end_t = times_v[(times_v > Ct_oi) & (fish_freq[id_idx] < Cblf + Crise_size * 0.37)] if len(rise_end_t) == 0: right_time_bound[enu] = np.nan else: right_time_bound[enu] = rise_end_t[0] mask = (~np.isnan(right_time_bound) & ((right_time_bound - left_time_bound) > 1.)) left_time_bound = left_time_bound[mask] right_time_bound = right_time_bound[mask] lower_freq_bound = lower_freq_bound[mask] upper_freq_bound = upper_freq_bound[mask] left_time_bound -= 0.01 * (t1 - t0) right_time_bound += 0.05 * (t1 - t0) lower_freq_bound -= 0.01 * (f1 - f0) upper_freq_bound += 0.05 * (f1 - f0) mask2 = ((left_time_bound >= t0) & (right_time_bound <= t1) & (lower_freq_bound >= f0) & (upper_freq_bound <= f1) ) left_time_bound = left_time_bound[mask2] right_time_bound = right_time_bound[mask2] lower_freq_bound = lower_freq_bound[mask2] upper_freq_bound = upper_freq_bound[mask2] x0 = np.array((left_time_bound - t0) / (t1 - t0) * width, dtype=int) x1 = np.array((right_time_bound - t0) / (t1 - t0) * width, dtype=int) y0 = np.array((1 - (upper_freq_bound - f0) / (f1 - f0)) * height, dtype=int) y1 = np.array((1 - (lower_freq_bound - f0) / (f1 - f0)) * height, dtype=int) bbox = np.array([[pic_save_str for i in range(len(left_time_bound))], left_time_bound, right_time_bound, lower_freq_bound, upper_freq_bound, x0, y0, x1, y1]) tmp_df = pd.DataFrame( data=bbox.T, columns=cols ) bbox_df = pd.concat([bbox_df, tmp_df], ignore_index=True) return bbox_df def main(args): # Hyperparameter min_freq = 200 max_freq = 1500 d_freq = 200 freq_overlap = 25 d_time = 60*10 time_overlap = 60*1 folders = list(f.parent for f in Path(args.folder).rglob('fill_times.npy')) if not args.inference: print('generate training dataset only for files with detected rises') folders = [folder for folder in folders if (folder / 'analysis' / 'rise_idx.npy').exists()] cols = ['image', 't0', 't1', 'f0', 'f1', 'x0', 'y0', 'x1', 'y1'] bbox_df = pd.DataFrame(columns=cols) # ToDo: Implement loading the old .csv file and upgrade it... how exactly I will determine after my vaccation # bbox_df = pd.read_csv(os.path.join('dataset', 'bbox_dataset.csv'), sep=',', index_col=0) # cols = list(bbox_df.keys()) # eval_files = [] # for f in pd.unique(bbox_df['image']): # eval_files.append(f.split('__')[0]) else: print('generate inference dataset ... only image output') bbox_df = {} for enu, folder in enumerate(folders): print(f'DataSet generation from {folder} | {enu+1}/{len(folders)}') # load different categories of data freq, times, spec = ( load_spec_data(folder)) EODf_v, ident_v, idx_v, times_v = ( load_tracking_data(folder)) if not args.inference: fish_freq, rise_idx, rise_size, fish_baseline_freq, fish_baseline_freq_time = ( load_trial_data(folder)) # generate iterator for analysis window loop pic_base = tqdm(itertools.product( np.arange(0, times[-1], d_time), np.arange(min_freq, max_freq, d_freq) ), total=int((((max_freq-min_freq)//d_freq)+1) * ((times[-1] // d_time)+1)) ) for t0, f0 in pic_base: t1 = t0 + d_time + time_overlap f1 = f0 + d_freq + freq_overlap present_freqs = EODf_v[(~np.isnan(ident_v)) & (t0 <= times_v[idx_v]) & (times_v[idx_v] <= t1) & (EODf_v >= f0) & (EODf_v <= f1)] if len(present_freqs) == 0: continue # get spec_idx for current spec snippet f_idx0, f_idx1 = np.argmin(np.abs(freq - f0)), np.argmin(np.abs(freq - f1)) t_idx0, t_idx1 = np.argmin(np.abs(times - t0)), np.argmin(np.abs(times - t1)) # get spec snippet and create torch.tensfor from it s = torch.from_numpy(spec[f_idx0:f_idx1, t_idx0:t_idx1].copy()).type(torch.float32) log_s = torch.log10(s) transformed = T.Normalize(mean=torch.mean(log_s), std=torch.std(log_s)) s_trans = transformed(log_s.unsqueeze(0)) pic_save_str, (width, height) = save_spec_pic(folder, s_trans, times, freq, t_idx0, t_idx1, f_idx0, f_idx1, args.dataset_folder) if not args.inference: bbox_df = bboxes_from_file(times_v, fish_freq, rise_idx, rise_size, fish_baseline_freq_time, fish_baseline_freq, pic_save_str, bbox_df, cols, width, height, t0, t1, f0, f1) if not args.inference: print('save bboxes') bbox_df.to_csv(os.path.join(args.dataset_folder, 'bbox_dataset.csv'), columns=cols, sep=',') if __name__ == '__main__': parser = argparse.ArgumentParser(description='Evaluated electrode array recordings with multiple fish.') parser.add_argument('folder', type=str, help='single recording analysis', default='') parser.add_argument('-d', "--dataset_folder", type=str, help='designated datasef folder', default='dataset') parser.add_argument('-i', "--inference", action="store_true", help="generate inference dataset. Img only") args = parser.parse_args() if not Path(args.dataset_folder).exists(): Path(args.dataset_folder).mkdir(parents=True, exist_ok=True) main(args)