starfys / timechange Goto Github PK

Closing the gui and reloadig the default project does read the files again so that part works, we just arent triggering it on adding files.

• workaround close the gui and reload the project

Using TensorFlow backend.
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\remof\AppData\Local\Programs\Python\Python35\lib\tkinter_init_.py", line 1550, in call
return self.func(*args)
File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange\gui.py", line 131, in genFFT
t = Thread(target=self.parent.tc.convert_all_csv, args=(self.parent.ConfigureScreen.method, self.parent.ConfigureScreen.chunksize, self.parent.ConfigureScreen.fftsize))
AttributeError: 'ConfigureScreen' object has no attribute 'method'
note - chandu is fixin right now

Multiple files may be selected
Must

pretty sure it was working before...

• workaround manually remove columns from csv files and reload gui app

The system should store the model in some way. Keras default is h5, so we may use this
Must

I thought steven already fixed this on sunday!
the same multithreading problem we discussed before?

• workaround this is fatal and there is no workaround
  Exception in Tkinter callback
  Traceback (most recent call last):
  File "C:\Users\remof\AppData\Local\Programs\Python\Python35\lib\tkinter_init_.py", line 1550, in call
  return self.func(*args)
  File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange\gui.py", line 278, in train_in_thread
  training_results = self.parent.tc.train()
  File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange_init_.py", line 287, in train
  if self.model is None:
  AttributeError: 'TimeChange' object has no attribute 'model'

leaves user entirely without instruction

• workaround be steven and already know what the right numbers are
  also, this is in teh wrong place, was on the configure classifier tab but that is properly the classifier config file with save button, this needs improved instructions an dmoved to its own tab or something

The system should store the generated images in a directory local to the software application or one specified by the user. Note, you may want to include the label in the name of the file.
Must

maybe this is becasue generating fft crashes silently without any output to the console window or maybe it finishes but never updates the cursor. unsure!

• workaround unsure this may actually be a fatal error (perhaps faith unwarrented)
  I think Chandu had a plan to fix this with a progress bar...

Using TensorFlow backend.
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "BestSplits" device_type: "CPU"') for unknown op: BestSplits
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "CountExtremelyRandomStats" device_type: "CPU"') for unknown op: CountExtremelyRandomStats
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "FinishedNodes" device_type: "CPU"') for unknown op: FinishedNodes
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "GrowTree" device_type: "CPU"') for unknown op: GrowTree
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "ReinterpretStringToFloat" device_type: "CPU"') for unknown op: ReinterpretStringToFloat
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "SampleInputs" device_type: "CPU"') for unknown op: SampleInputs
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "ScatterAddNdim" device_type: "CPU"') for unknown op: ScatterAddNdim
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "TopNInsert" device_type: "CPU"') for unknown op: TopNInsert
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "TopNRemove" device_type: "CPU"') for unknown op: TopNRemove
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "TreePredictions" device_type: "CPU"') for unknown op: TreePredictions
E c:\tf_jenkins\home\workspace\release-win\device\cpu\os\windows\tensorflow\core\framework\op_kernel.cc:943] OpKernel ('op: "UpdateFertileSlots" device_type: "CPU"') for unknown op: UpdateFertileSlots

[2:03]
on a similar note, who the hell is jenkins

[2:04]
https://www.youtube.com/watch?v=hooKVstzbz0
YouTube DBlow2003
Leeroy Jenkins HD (High Quality)

[2:06]
tensorflow/tensorflow#7500 apparently the jenkins bug is a bug in tensorflow itself so not somethign we can fix
GitHub
TensorFlow version 1.0.0-rc2 on Windows: “OpKernel ('op: ”BestSplits“ device_type: ”CPU“') for unknown op: BestSplits” with test code · Issue #7500 · tensorflow/tensorflow
I installed TensorFlow version 1.0.0-rc2 on Windows 7 SP1 x64 Ultimate (Python 3.5.2 |Anaconda custom (64-bit)) using: pip install --upgrade https://storage.googleapis.com/tensorflow/windows/cpu/te...

[2:08]
also it doesnt seem to be a "crash" per-se. if you ignore those messages and go train anyways it seems to be doing its thing

GUI must be allow users to include, exclude, and reorder columns for a CSV file.
Want

The conversion in FR03a should be done with additional methods, for example, wavelet
Want

Timechange needs installation instructions for

• Linux (debian)
• Max OS X
• Windows?

Gui runs in separate thread from operations like file i/o, training, and conversion.
Must

Timechange needs a basic set of instructions for normal usage

The system should provide a button to commence model training
Must

The system should offer users ability to specify parameters needed for conversion in FR03a. For example, information about desired window length, overlap ratio, and sampling rate
Must

Input file format: comma-separated value (.csv)
Must

FR05a
Must

Instead of writing test48.png, test49.png, test50.png it would be better if we preserved the labeling of the data when such a label is provided such as the directory name. Perhaps just putting the output in a directory of the same name as the input directory would suffice.

FR03a
Must

The system should allow conversion of the data to images with no transformations applied, just raw data
Want

The system should provide a button to commence the conversion process
Must

• workaround delete old project directory
  Using TensorFlow backend.
  Exception in Tkinter callback
  Traceback (most recent call last):
  File "C:\Users\remof\AppData\Local\Programs\Python\Python35\lib\tkinter_init_.py", line 1550, in call
  return self.func(*args)
  File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange\gui.py", line 29, in defaultProject
  self.parent.tc = timechange.TimeChange()
  File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange_init_.py", line 129, in init
  path.join(self.project_path, "parameters.conf")))
  Exception: C:\Users\remof\timechange\default cannot be a timechange project because C:\Users\remof\timechange\default\parameters.conf does not exist

The conversion in FR03a should be done based on FFT
Must

The system should generate a notification when the conversion process is completed
Must

a popup telling the user that stuff is happening would be "reassuring". Clicking on the fft preview tab tells you "work in progress" which is further misleading

• workaround faith

FR01a
Must

The system should store a reference to the supplied files rather than loading them into memory.
Must

The system should allow splitting the time series data into chunks of equal size.
Want

Labels for each file may come from either the file’s name or from the user via a GUI widget
Must

FR05E
The system should calculate and present—in a nice way—measures of accuracy of the trained model. I recommend 10-fold cross validation.
Must

Using TensorFlow backend.
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\remof\AppData\Local\Programs\Python\Python35\lib\tkinter_init_.py", line 1550, in call
return self.func(*args)
File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange\gui.py", line 279, in train_in_thread
training_results = self.parent.tc.train()
File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange_init_.py", line 287, in train
if self.model is None:
AttributeError: 'TimeChange' object has no attribute 'model'
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\Users\remof\AppData\Local\Programs\Python\Python35\lib\tkinter_init_.py", line 1550, in call
return self.func(*args)
File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange\gui.py", line 238, in save
self.parent.tc.build_model()
File "C:\Users\remof\Desktop\timechange-demo\timechange\timechange_init_.py", line 268, in build_model
if self.image_size is None:
AttributeError: 'TimeChange' object has no attribute 'image_size'

Perhaps a sucessful popup or switchign to the next task tab would be better behavior

• workaround verify function by looking for new files in directory

We need to be able to take multiple simultaneous time series data points, perform FFT on all of them, and concatenate their FFT signatures into one image.

• workaround resize manually with mouse

To do the training specified in FR05a, users must be able to enter training parameters for CNN models. # layers, etc
Must