Datasets:

JasonXF
/

SeismicTransformerData

Modalities:

Text

License:

Dataset card Data Studio Files Files and versions

xet

Community

SeismicTransformerData / MDOF /knet-250 /PythonScripts /5_combine_h5.py

JasonXF

Add files using upload-large-folder tool

787163a verified 3 months ago

Raw History Blame Contribute Delete

2.39 kB

import os
import re
import h5py
import numpy as np

def read_h5_data(h5_path, dataset_name):
    with h5py.File(h5_path, 'r') as h5_file:
        return h5_file[dataset_name][()]

def main(file_name:str):
    target_directory = os.getcwd()
    folder_pattern = re.compile(r'^(\d+)-(\d+)$')  # 修改正则表达式以匹配文件夹
    file_pattern = re.compile(r'^(\d+)-(\d+)\.h5$')  # 正则表达式匹配.h5文件

    print(f"target_directory = {target_directory}")
    # 获取所有匹配文件夹，并根据数字范围对它们进行排序
    folders = [f for f in os.listdir(target_directory) if folder_pattern.match(f)]
    folders.sort(key=lambda x: int(folder_pattern.match(x).group(1)))

    print(f"Found {len(folders)} folders: {folders}")

    # 初始化列表以存储拼接后的数组
    Acc_Floor_Responses = []
    DS_Blgs = []

    # 循环读取每个文件夹中的HDF5文件，并拼接数据
    for folder in folders:
        h5_file_name = f"{folder}.h5"
        h5_path = os.path.join(target_directory, folder, h5_file_name)  # 文件路径包括文件夹名称
        if os.path.exists(h5_path) and file_pattern.match(h5_file_name):
            print(f"Processing {h5_path}...")
            Acc_Floor_Response = read_h5_data(h5_path, 'Acc_Floor_Response')
            DS_Blg = read_h5_data(h5_path, 'DS_Blg')

            # 将数据添加到列表中
            Acc_Floor_Responses.append(Acc_Floor_Response)
            DS_Blgs.append(DS_Blg)
        else:
            print(f"Expected file {h5_file_name} not found in {folder}")

    # 使用numpy.concatenate进行数组拼接
    Acc_Floor_Responses = np.concatenate(Acc_Floor_Responses, axis=0)
    DS_Blgs = np.concatenate(DS_Blgs, axis=0)

    # 打印出拼接后的数组形状以检查
    print(f"Acc_Floor_Responses shape: {Acc_Floor_Responses.shape}")
    print(f"DS_Blgs shape: {DS_Blgs.shape}")

    # 可以选择保存拼接后的大数组
    h5_combined_path = os.path.join(target_directory, file_name)
    with h5py.File(h5_combined_path, 'w') as h5_combined:
        h5_combined.create_dataset('Acc_Floor_Response', data=Acc_Floor_Responses)
        h5_combined.create_dataset('Blg_Damage_State', data=DS_Blgs)
    print(f"Combined data saved to {h5_combined_path}")

if __name__ == "__main__":
    main('Blg_F2_6m_IM7_SCD2.h5')

	import os
	import re
	import h5py
	import numpy as np

	def read_h5_data(h5_path, dataset_name):
	with h5py.File(h5_path, 'r') as h5_file:
	return h5_file[dataset_name][()]

	def main(file_name:str):
	target_directory = os.getcwd()
	folder_pattern = re.compile(r'^(\d+)-(\d+)$') # 修改正则表达式以匹配文件夹
	file_pattern = re.compile(r'^(\d+)-(\d+)\.h5$') # 正则表达式匹配.h5文件

	print(f"target_directory = {target_directory}")
	# 获取所有匹配文件夹，并根据数字范围对它们进行排序
	folders = [f for f in os.listdir(target_directory) if folder_pattern.match(f)]
	folders.sort(key=lambda x: int(folder_pattern.match(x).group(1)))

	print(f"Found {len(folders)} folders: {folders}")

	# 初始化列表以存储拼接后的数组
	Acc_Floor_Responses = []
	DS_Blgs = []

	# 循环读取每个文件夹中的HDF5文件，并拼接数据
	for folder in folders:
	h5_file_name = f"{folder}.h5"
	h5_path = os.path.join(target_directory, folder, h5_file_name) # 文件路径包括文件夹名称
	if os.path.exists(h5_path) and file_pattern.match(h5_file_name):
	print(f"Processing {h5_path}...")
	Acc_Floor_Response = read_h5_data(h5_path, 'Acc_Floor_Response')
	DS_Blg = read_h5_data(h5_path, 'DS_Blg')

	# 将数据添加到列表中
	Acc_Floor_Responses.append(Acc_Floor_Response)
	DS_Blgs.append(DS_Blg)
	else:
	print(f"Expected file {h5_file_name} not found in {folder}")

	# 使用numpy.concatenate进行数组拼接
	Acc_Floor_Responses = np.concatenate(Acc_Floor_Responses, axis=0)
	DS_Blgs = np.concatenate(DS_Blgs, axis=0)

	# 打印出拼接后的数组形状以检查
	print(f"Acc_Floor_Responses shape: {Acc_Floor_Responses.shape}")
	print(f"DS_Blgs shape: {DS_Blgs.shape}")

	# 可以选择保存拼接后的大数组
	h5_combined_path = os.path.join(target_directory, file_name)
	with h5py.File(h5_combined_path, 'w') as h5_combined:
	h5_combined.create_dataset('Acc_Floor_Response', data=Acc_Floor_Responses)
	h5_combined.create_dataset('Blg_Damage_State', data=DS_Blgs)
	print(f"Combined data saved to {h5_combined_path}")

	if __name__ == "__main__":
	main('Blg_F2_6m_IM7_SCD2.h5')