metrofun / e3d-lstm Goto Github PK

Download data from OneDrive or BaiduYun is lost
I need the data to test
Please you issue it by the new link
thank you !

I wonder why the mse value in the code is multiplied by 255

Link to notebook
Click here

Steps to reproduce this issue:
1.cd /content/E3D-LSTM/src/
2.sys.path.append('/content/E3D-LSTM/src')
3. !python -m trainer

Error Stack Trace
Traceback (most recent call last): File "/usr/lib/python3.6/runpy.py", line 193, in _run_module_as_main "__main__", mod_spec) File "/usr/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/content/E3D-LSTM/src/trainer.py", line 177, in <module> trainer.resume_train() File "/content/E3D-LSTM/src/trainer.py", line 132, in resume_train train_dataloader = self.get_trainloader(self.data[:-672]) File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 576, in __getattr__ type(self).__name__, name)) AttributeError: 'TaxiBJTrainer' object has no attribute 'data'

Hi. The code you made seems very useful to me.

I tried to train E3D-LSTM and save TaxiBJ result data, but I couldn't save because there are two channels. How can I visualize TaxiBJ data??
I would really appreciate it if you could give me some code or functions to reference. Thank you.

Dataset failure，Can you share it again？Thanks.

Hey there!

Thanks for the library. I was wondering if I could get some insight into what these hyperparameters "are"? I've downloaded the dataset and played around with it setting breakpoints, but the names are confusing me.

        self.input_time_window = 4
        self.output_time_horizon = 1
        self.temporal_stride = 1
        self.temporal_frames = 2
        self.time_steps = (
            self.input_time_window - self.temporal_frames + 1
        ) // self.temporal_stride

input_time_window

I'm not sure what this is....

output_time_horizon

I'm guessing that output_time_horizon is the length of the prediction? E.g if output_time_horizon == 1, output a scalar, if output_time_horizon == 2 output a vector of length 2?

temporal stride

Given some input timeseries of [1,2,3,4,5,6, 7] with stride 2 and size 3

T1 = [1,2,3]
T2 = [3,4,5]

?

temporal_frames

Is this the window size as mentioned above?

The error "torch.nn.modules.module.ModuleAttributeError: 'TaxiBJTrainer' object has no attribute 'seek'" occurs when I set the resume=True. I change the code in trainer.py
if resume:
checkpoint = torch.load(self, ckpt_path)
epoch = checkpoint["epoch"]
model.load_state_dict(checkpoint["state_dict"])
to
if resume:
checkpoint = torch.load(ckpt_path)
epoch = checkpoint["epoch"]
self.load_state_dict(checkpoint["state_dict"])
which fix the error.