using Plotting=true `FileNotFoundError: [Errno 2] No such file or directory: `

[jonahpc]

Is there an existing issue for this?

• I have searched the existing issues

Bug description

when I let evalute_network(config_path, plotting=True) I get an error. This is not true when I have plot=False.

Operating System

Windows 10

DeepLabCut version

2.2.0.2

DeepLabCut mode

single animal

Device type

Geforce RTX 2080t

Steps To Reproduce

No response

Relevant log output

Running  DLC_resnet50_Right_Side_cameraOct28shuffle1_50000  with # of trainingiterations: 50000
C:\Users\Nischal\Anaconda3\envs\dlc2\lib\site-packages\tensorflow\python\keras\engine\base_layer_v1.py:1694: UserWarning: `layer.apply` is deprecated and will be removed in a future version. Please use `layer.__call__` method instead.
  warnings.warn('`layer.apply` is deprecated and '
Analyzing data...
42it [00:05,  7.30it/s]
Done and results stored for snapshot:  snapshot-50000
Results for 50000  training iterations: 95 1 train error: 2.67 pixels. Test error: 4.14  pixels.
With pcutoff of 0.6  train error: 2.67 pixels. Test error: 4.14 pixels
Thereby, the errors are given by the average distances between the labels by DLC and the scorer.
Plotting...
 43%|███████████████████████████████████▏                                              | 18/42 [00:05<00:06,  3.49it/s]
---------------------------------------------------------------------------
FileNotFoundError                         Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_14292/4290122146.py in <module>
----> 1 deeplabcut.evaluate_network(path_config_file, plotting=True)

~\Anaconda3\envs\dlc2\lib\site-packages\deeplabcut\pose_estimation_tensorflow\core\evaluate.py in evaluate_network(config, Shuffles, trainingsetindex, plotting, show_errors, comparisonbodyparts, gputouse, rescale, modelprefix)
    896                             )
    897                             auxiliaryfunctions.attempttomakefolder(foldername)
--> 898                             Plotting(
    899                                 cfg,
    900                                 comparisonbodyparts,

~\Anaconda3\envs\dlc2\lib\site-packages\deeplabcut\pose_estimation_tensorflow\core\evaluate.py in Plotting(cfg, comparisonbodyparts, DLCscorer, trainIndices, DataCombined, foldername)
    199     fig, ax = visualization.create_minimal_figure()
    200     for ind in tqdm(np.arange(NumFrames)):
--> 201         ax = visualization.plot_and_save_labeled_frame(
    202             DataCombined,
    203             ind,

~\Anaconda3\envs\dlc2\lib\site-packages\deeplabcut\utils\visualization.py in plot_and_save_labeled_frame(DataCombined, ind, trainIndices, cfg, colors, comparisonbodyparts, DLCscorer, foldername, fig, ax, scaling)
    172         ax=ax,
    173     )
--> 174     save_labeled_frame(fig, image_path, foldername, ind in trainIndices)
    175     return ax
    176 

~\Anaconda3\envs\dlc2\lib\site-packages\deeplabcut\utils\visualization.py in save_labeled_frame(fig, image_path, dest_folder, belongs_to_train)
    191         full_path = "\\\\?\\" + full_path
    192     fig.subplots_adjust(left=0, bottom=0, right=1, top=1, wspace=0, hspace=0)
--> 193     fig.savefig(full_path)
    194 
    195 

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\figure.py in savefig(self, fname, transparent, **kwargs)
   3013                 patch.set_edgecolor('none')
   3014 
-> 3015         self.canvas.print_figure(fname, **kwargs)
   3016 
   3017         if transparent:

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\backends\backend_wx.py in print_figure(self, filename, *args, **kwargs)
    645     def print_figure(self, filename, *args, **kwargs):
    646         # docstring inherited
--> 647         super().print_figure(filename, *args, **kwargs)
    648         # Restore the current view; this is needed because the artist contains
    649         # methods rely on particular attributes of the rendered figure for

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\backend_bases.py in print_figure(self, filename, dpi, facecolor, edgecolor, orientation, format, bbox_inches, pad_inches, bbox_extra_artists, backend, **kwargs)
   2253                 # force the figure dpi to 72), so we need to set it again here.
   2254                 with cbook._setattr_cm(self.figure, dpi=dpi):
-> 2255                     result = print_method(
   2256                         filename,
   2257                         facecolor=facecolor,

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\backend_bases.py in wrapper(*args, **kwargs)
   1667             kwargs.pop(arg)
   1668 
-> 1669         return func(*args, **kwargs)
   1670 
   1671     return wrapper

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\backends\backend_agg.py in print_png(self, filename_or_obj, metadata, pil_kwargs, *args)
    507         """
    508         FigureCanvasAgg.draw(self)
--> 509         mpl.image.imsave(
    510             filename_or_obj, self.buffer_rgba(), format="png", origin="upper",
    511             dpi=self.figure.dpi, metadata=metadata, pil_kwargs=pil_kwargs)

~\Anaconda3\envs\dlc2\lib\site-packages\matplotlib\image.py in imsave(fname, arr, vmin, vmax, cmap, format, origin, dpi, metadata, pil_kwargs)
   1614         pil_kwargs.setdefault("format", format)
   1615         pil_kwargs.setdefault("dpi", (dpi, dpi))
-> 1616         image.save(fname, **pil_kwargs)
   1617 
   1618 

~\Anaconda3\envs\dlc2\lib\site-packages\PIL\Image.py in save(self, fp, format, **params)
   2235                 fp = builtins.open(filename, "r+b")
   2236             else:
-> 2237                 fp = builtins.open(filename, "w+b")
   2238 
   2239         try:

FileNotFoundError: [Errno 2] No such file or directory: '\\\\?\\E:/DLC2-Working/Right_Side_camera-JPC-2021-10-28\\evaluation-results\\iteration-0\\Right_Side_cameraOct28-trainset95shuffle1\\LabeledImages_DLC_resnet50_Right_Side_cameraOct28shuffle1_50000_snapshot-50000\\Training-A56M2_100PerXy_10min_in_1_awake_Right-img05711.png'

Anything else?

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[jonahpc]

I think the issue has something to do with python 3.8... I am storing my videos on a NAS, which I have mapped to my "Z" drive. I've heard rumors that python 3.8 is cranky about network drives. Preemptively, thanks a ton for the help; it means the world!

[MMathisLab]

Hi @jonahpc seems a path issue:

No such file or directory: '\\\\?\\E:/DLC2-Working/Right_Side_camera-JPC-2021-10-28\\evaluation-results\\iteration-0\\Right_Side_cameraOct28-trainset95shuffle1\\LabeledImages_DLC_resnet50_Right_Side_cameraOct28shuffle1_50000_snapshot-50000\\Training-A56M2_100PerXy_10min_in_1_awake_Right-img05711.png'

maybe the image is not accessible; esp. see the \ and / mismatch

[jonahpc]

Hello @MMathisLab! Thanks a ton for the response!

So i have copied and pasted the path from my finder, and have corrected for backslashes. Matlab notices my network attached storage (NAS) path, as I have mapped it locally.

But, when I use:
"import os
print(os.path.realpath())"

I get that the file path does not exists. I think it has to do with the fact that the harddrive is on my network. DLC used to work with this set-up. Could it possibly be a new python thing? If so, have you any experience getting around this error?

Thanks @MMathisLab so much!

[MMathisLab]

Hi @jonahpc I am not sure / hard for me to test as I don't have this set up, sorry! '\\\\?\\E:/ But I think this is the issue; maybe check your config.yaml to be sure the video path link does not start with ?

[Guillermo-Hidalgo-Gadea]

Hi,
just experienced the same issue working from an external HDD, not NAS, with Windows, DeepLabCut 2.2.0.2 and python 3.8.12.
I noticed that my config.yaml containes a few ? for specific videos (actually those in h.265 but not those in h.264) and that the project_path: is formated in slashes, while the video_sets: are in backslashes.

What solved the error above for me was not changing the paths in the config.yaml file (these get overwritten) but changing the config_path in deeplabcut.evaluate_network(config_path) to backslashes.

This seems to have changed during the last updates (python 3.8?), since getting the config_path from tk with filedialog.askopenfilenames() worked fine before...

[MMathisLab]

oh indeed that might be a new python update; btw, this PR woudl finally solve this issue, if you want to test it out! #1584

[jeylau]

Closing as #1584 has now been merged and should solve the issue. Feel free to reopen it if that's not the case 😊