'''
import tensorflow as tf
import numpy as np
import os
import pandas as pd
from classifier_dataset import classifier_25, SaveFile, LoadFile, fft_transformer, guiyi
from re_sub import acc_regression
if __name__ == '__main__':
path = '/home/xiaosong/pny相关数据/data_pny/PNY_all.pickle'
path_cl = '/home/xiaosong/桌面/graph_cl_re/graph_cl.h5'
path_re = '/home/xiaosong/桌面/graph_cl_re/graph_re.h5'
model_cl = tf.keras.models.load_model(filepath=path_cl)
model_re = tf.keras.models.load_model(filepath=path_re)
space_list = classifier_25(n=26)
dataset = LoadFile(p=path)
dataset_4feature, dataset_dense, label = dataset[:, :4], dataset[:, 4:-1], \
dataset[:, -1][:, np.newaxis]
dataset_fft = fft_transformer(dataset_dense, 100)
dataset = np.hstack((dataset_4feature, dataset_fft, label))
dataset_guiyi = guiyi(dataset)
print(dataset_guiyi.shape)
rng = np.random.RandomState(0)
rng.shuffle(dataset_guiyi)
test_data = dataset_guiyi[:6000, :] #根据导入数据标签改
print(test_data.shape)
result_cl = model_cl.predict(x=[test_data[:, :4], test_data[:, 4:-1]],
verbose=0)
result_cl = np.argmax(a=result_cl, axis=1)
result_inf = result_cl * 10 #将分类器执行后的结果对应到各个子分类空间中
#regression
np.vstack((dataset_return, dataset_sub[:number, :]))
elif dataset_sub.shape[0] and dataset_sub.shape[0] < number:
judge = number % dataset_sub.shape[0]
num = number // dataset_sub.shape[0]
if judge != 0:
num += 1
dataset_sub2000 = dataset_sub
for i in range(num - 1):
dataset_sub2000 = np.vstack((dataset_sub2000, dataset_sub))
dataset_sub2000 = dataset_sub2000[:number, :]
dataset_return = dataset_sub2000 if dataset_return.any() == 0 else \
np.vstack((dataset_return, dataset_sub2000))
return dataset_return
if __name__ == '__main__':
space = classifier_25(26)
# print(space)
p = '/home/xiaosong/oldenburg相关数据/data_oldenburg/OLDENBURG_all.pickle'
dataset = LoadFile(p)
dataset_guiyi_sub = dataset_regression_guiyi(dataset, space, number=2000)
# print(dataset_guiyi_sub.shape)
dataset_4feature, dataset_dense, label = dataset_guiyi_sub[:, :4], dataset_guiyi_sub[:, 4:-1], \
dataset_guiyi_sub[:, -1][:, np.newaxis]
dataset_fft = fft_transformer(dataset_dense, 100)
dataset = np.hstack((dataset_4feature, dataset_fft, label))
dataset_guiyi = guiyi(dataset)
print(dataset_guiyi.shape)
SaveFile(data=dataset_guiyi,
savepickle_p='/home/xiaosong/桌面/oldenburg_regression_sub.pickle')
print(np.max(dataset_guiyi, axis=0))