search

santteegt / ucl-drl-msc

Learning to Recommend using a Deep Reinforcement Agent
Apache License 2.0
22 stars 7 forks source link

Exception gym.error.Error #2

Open 4575759ww opened 7 years ago

4575759ww commented 7 years ago

Hi again, I have downloaded your dataset successfully, thank you! But now, I get an error again ,hope you can help.

The log file is Exception gym.error.Error: Error('env has been garbage collected. To keep using a monitor, you must keep around a reference to the env object. (HINT: try assigning the env to a variable in your code.)',) in <bound method Monitor.del of <gym.monitoring.monitor.Monitor object at 0x11b50a3d0>> ignored

Look forward to your reply.

santteegt commented 7 years ago

Which version of python are you using ?

4575759ww commented 7 years ago

python 2.7

4575759ww commented 7 years ago

Hi @santteegt ,I have already successfully run the program，thank you！ I noticed that there is a file--u.complete.data.csv in your dataset, and the last column is prev-item. How did you calculate the prev-item？I find its value contains -1， what does this mean？

santteegt commented 7 years ago

Hello,

Good to know that you could solved that issue. What was the problem?.

-1 means that it is the 1st recommended item for that user.

Regards

4575759ww commented 7 years ago

Thank you for your reply! The reason was about openai gym installing , something I missed. I should follow your steps.

I almost guessed the meaning of -1, so you mean the prev-item is the recommended item? But in your code annotation, you say the prev-item means state, the item means action, so my opinion about the results in the last run.log file, the prev-item means the movie which user have seen or be seeing, and item means the recommended movie, have I misunderstood your meaning? Hope you can help me to understand it.

Also, I read the transition_matrix.py file, and I want to run it , could you please tell me the contents of raw_data or give it directly？ Thank you

Hope you well！

santteegt commented 7 years ago

What you're saying is right. I just tried to explain that a register with prev-item equals to -1 means that it is the 1st recommended item for that user, so no previous items were recommended to him/her.

About running the transition_matrix.py. I'm a little busy right now, but I'm going to try to give me time space to review that code and give you and answer.

Regards

4575759ww commented 7 years ago

OK，I am much appreciated that you will review ,thank you for your answer, it helps a lot.

kangyang73 commented 6 years ago

I encountered the same problem as you, can you tell me how you solved it?

kangyang73 commented 6 years ago

@4575759ww I encountered the same problem as you, can you tell me how you solved it?

Look forward to your reply.

santteegt commented 6 years ago

Hello,

I'm really sorry for not replying to your messages for so long (busy months in my new research position). I'm going to continue with this research topic in the following weeks, and I hope I can give an immediate solution to your issue as soon as possible.

Regards