data_process_example/combine_csi_cv.py

import pickle
import numpy as np

with open("./data_sequence.pkl", 'rb') as f:
# with open("./data_sequence_linear.pkl", 'rb') as f:
    data_csi = pickle.load(f)
with open("./gt_data.pkl", 'rb') as f:
    data_cv = pickle.load(f)

data = []
pad = -1000
for k in range(len(data_csi)):
    csi = data_csi[k]
    cv = data_cv[k]

    x = cv['x']
    y = cv['y']
    img_path = np.array(cv['img_path'])
    time_cv = cv['timestamp']
    print(cv['people_name'])

    indices = np.argsort(time_cv)
    x = x[indices]
    y = y[indices]
    img_path = img_path[indices]
    time_cv = time_cv[indices]

    print(x)
    print(y)


    csi_time = csi['global_time']
    local_time = csi['time']
    magnitude = csi['magnitude']
    phase = csi['phase']
    people = csi['people']

    indices = np.argsort(csi_time)
    local_time = local_time[indices]
    magnitude = magnitude[indices]
    csi_time = csi_time[indices]
    phase = phase[indices]



    x_list = []
    y_list = []
    path_list = []

    i = 0
    j = 0

    print(csi_time)
    print(time_cv)


    while csi_time[i] < time_cv[j]:
        i += 1
        x_list.append(pad)
        y_list.append(pad)
        path_list.append(pad)


    # print(len(csi_time))
    # print(len(time_cv))
    while i < len(csi_time):
        while csi_time[i] > time_cv[j]:
            j += 1
            if j >= len(time_cv):
                break
        if j >= len(time_cv):
            break
        x_list.append(x[j])
        y_list.append(y[j])
        path_list.append(img_path[j])
        i += 1

    print(len(x_list))


    if len(x_list) < len(csi_time):
        num = len(csi_time) - len(x_list)
        x_list = x_list + [pad] * num
        y_list = y_list + [pad] * num
        path_list = path_list + [pad] * num

    data.append({
        'magnitude': magnitude,
        'phase': phase,
        'x': x_list,
        'y': y_list,
        'img_path': path_list,
        'time': local_time,
        'people': people
    })
    print(people)
    print(len(magnitude),len(phase),len(x_list),len(y_list),len(path_list),len(local_time))

output_file = './wiloc.pkl'
# output_file = './wiloc_linear.pkl'
with open(output_file, 'wb') as f:
    pickle.dump(data, f)