2021-06-22

DeepLabCut 学习指南1

- 安装环境DLC-CPU版本，系统环境Windows

• 安装参考官方给的视频以及下方GitHub链接

• 需要注意的是，视频中使用的git clone 下载方法本人没成功实现，所以直接进入GitHub下载zip安装包

• 根据视频的步骤操作基本上安装就没什么问题

- 安装遇到的问题

• 1-

• 遇到FutureWarning: Passing (type, 1) or ‘1type’ as a synonym of type is deprecate
  

• 出现这种情况我查看了我numpy的版本为1.17.5（通过DLC-CPU.yaml文件直接装载的虚拟环境）

• 解决参考解决python调用TensorFlow时出现FutureWarning: Passing (type, 1) or ‘1type’ as a synonym of type is deprecate文章的方法将numpy版本从1.17.5降级到了1.16.5

• 2-

• 在实现创建项目，标好label，准备create trianing dataset 遇到问题End of HDF5 error back trace，然后到现在我也无法解决。

`Traceback (most recent call last):
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\deeplabcut\gui\create_training_dataset.py", line 391, in create_training_dataset
    augmenter_type=self.aug_choice.GetValue(),
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\deeplabcut\generate_training_dataset\trainingsetmanipulation.py", line 882, in create_training_dataset
    cfg, Path(os.path.join(project_path, trainingsetfolder)), windows2linux
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\deeplabcut\generate_training_dataset\trainingsetmanipulation.py", line 634, in merge_annotateddatasets
    AnnotationData.to_hdf(filename + ".h5", key="df_with_missing", mode="w")
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\pandas\core\generic.py", line 2621, in to_hdf
    encoding=encoding,
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\pandas\io\pytables.py", line 278, in to_hdf
    path_or_buf, mode=mode, complevel=complevel, complib=complib
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\pandas\io\pytables.py", line 561, in __init__
    self.open(mode=mode, **kwargs)
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\pandas\io\pytables.py", line 710, in open
    self._handle = tables.open_file(self._path, self._mode, **kwargs)
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\tables\file.py", line 315, in open_file
    return File(filename, mode, title, root_uep, filters, **kwargs)
  File "D:\Anaconda\anaconda\envs\DLC-CPU\lib\site-packages\tables\file.py", line 778, in __init__
    self._g_new(filename, mode, **params)
  File "tables\hdf5extension.pyx", line 492, in tables.hdf5extension.File._g_new
tables.exceptions.HDF5ExtError: HDF5 error back trace

  File "D:\pytables_hdf5\CMake-hdf5-1.10.5\hdf5-1.10.5\src\H5F.c", line 444, in H5Fcreate
    unable to create file
  File "D:\pytables_hdf5\CMake-hdf5-1.10.5\hdf5-1.10.5\src\H5Fint.c", line 1509, in H5F_open
    unable to open file: time = Tue Jun 22 15:08:39 2021
, name = 'F:\deeplabcut\test-fj-2021-06-18\training-datasets\iteration-0\UnaugmentedDataSet_test六月18\CollectedData_fj.h5', tent_flags = 13
  File "D:\pytables_hdf5\CMake-hdf5-1.10.5\hdf5-1.10.5\src\H5FD.c", line 734, in H5FD_open
    open failed
  File "D:\pytables_hdf5\CMake-hdf5-1.10.5\hdf5-1.10.5\src\H5FDsec2.c", line 346, in H5FD_sec2_open
    unable to open file: name = 'F:\deeplabcut\test-fj-2021-06-18\training-datasets\iteration-0\UnaugmentedDataSet_test 六月18\CollectedData_fj.h5', errno = 2, error message = 'No such file or directory', flags = 13, o_flags = 302

End of HDF5 error back trace

Unable to open/create file 'F:\deeplabcut\test-fj-2021-06-18\training-datasets\iteration-0\UnaugmentedDataSet_test六月18\CollectedData_fj.h5'`

• 21-6-23更新，我一直觉得是因为中文日期的出现导致了问题。但是没有足个File去对，然后就查“End of HDF5 error back trace”，结果没有找到。
• 尝试将系统日期改为英文版本，失败。
• 然后逐个File溯源。在路径下找到了py文件然后根据报错位置查看。
  
  
• 在GetTrainingSetFolder里面的date获取就应该是英文日期，但是项目的yaml文件中date还是中文的，所以把中文的date改为英文，done！
  
  

- DeepLabCut GUI 操作指南

• 第一条建议！very important！请使用GPU！！！
  -本人一开始使用CPU跑默认设置的Resnet50，一个晚上iteration才到1000，默认参数设置如下。目前准备换服务器跑，看来是我年少无知了。GPU版本的Deeplabcut安装指南后续出。