wuhuikai / msc Goto Github PK
View Code? Open in Web Editor NEWMSC: A Dataset for Macro-Management in StarCraft II
MSC: A Dataset for Macro-Management in StarCraft II
Hi, thanks for making this repo and release your macro dataset. In your dataset for example Terran, the action space is limited to 75 actions, 0-74. How can I map it back to pysc2's original action space? (0-523)
Thank you so much
Hi,
To read the action from the replays, is there a way to read the action arguments as well? For now it's just the action ID and no arguments.
Thanks
I installed a stream module with pip but there is no function 'open' in this module.
with open('a.SC2Replay') as f:
json.load(f)
UnicodeDecodeError: 'gbk' codec can't decode byte 0xac in position 60: illegal multibyte sequence
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb0 in position 60: invalid start byte
I tried many times, but still fail to read the replay in json
when I run extract_actions.py,it print this error :"Connection already closed. SC2 probably crashed. Check the error log.". And then the sc2 client shows a dialog which means unable to open replay.Do you know something about it?
File "replay_stat.py", line 124, in main
update(replay, stat)
File "replay_stat.py", line 33, in update
states = json.load(f)
File "/home/jh/anaconda3/envs/pymarl/lib/python3.6/json/init.py", line 299, in load
parse_constant=parse_constant, object_pairs_hook=object_pairs_hook, **kw)
File "/home/jh/anaconda3/envs/pymarl/lib/python3.6/json/init.py", line 354, in loads
return _default_decoder.decode(s)
File "/home/jh/anaconda3/envs/pymarl/lib/python3.6/json/decoder.py", line 339, in decode
obj, end = self.raw_decode(s, idx=_w(s, 0).end())
File "/home/jh/anaconda3/envs/pymarl/lib/python3.6/json/decoder.py", line 355, in raw_decode
obj, end = self.scan_once(s, idx)
json.decoder.JSONDecodeError: Expecting ':' delimiter: line 1 column 695279 (char 695278)
#SET: 33%|███████████████████████████████████████▎ | 1/3 [00:59<01:58, 59.20s/it]
#Replay: 30%|████████████████████████████████▋ | 2947/9754 [00:59<02:16, 49.80it/s]
Recreating the database with a min_mrr flag of 4000, and there is a runtime error in parse_replay_info.py
You need to call FLAGS(sys.argv) after the flags.DEFINE methods.
Hi,
First, thanks for making this repository.
And I'm trying to go through the "hard way" to get actions with all details(the parameters), and when I run python preprocess.py
I get the error as
xuan@blairlab:/datasets/MSC/preprocess$ python3 preprocess.py
Version: B55958 (SC2.3.16)
Build: Jul 31 2017 13:19:41
Command Line: '"/home/xuan/StarCraftII/Versions/Base55958/SC2_x64" -listen 127.0.0.1 -port 20499 -dataDir /home/xuan/StarCraftII/ -tempDir /tmp/sc-zvqnp6rt/ -displayMode 0'
Starting up...
Startup Phase 1 complete
Startup Phase 2 complete
Creating stub renderer...
Listening on: 127.0.0.1:20499 (20499)
Startup Phase 3 complete. Ready for commands.
W0321 21:03:31.184315 140275951630080 sc_process.py:183] Killing the process.
#Replay: 0%| | 0/64395 [00:00<?, ?it/s]Traceback (most recent call last):
File "preprocess.py", line 89, in <module>
main()
File "preprocess.py", line 77, in main
for player_info in players_info))
File "preprocess.py", line 77, in <genexpr>
for player_info in players_info))
AttributeError: module 's2clientprotocol.sc2api_pb2' has no attribute 'Race'
I'm using the 3.16.1 version of the game for linux btw.
Could you please have a look on this when you get time, thanks.
Is there a mechanism\module\function with which we can extract the names of the columns from the GlobalFeatures and probably transform it to a single dataframe?
I've download your global features vector consist of lots of files postfix ".npz". There are 5 subdatasets in each file when I load them and I can't understand it with your describe of "Global Features Vector". It should be a matrix F(T,M) as you say, but I can't correspond which I download with the matrix F(T,M).
I'm a graduate student. Looking forward to your reply.
Hi,
When I try to parse replay info using the given file parse_replay_info.py I get the error seen below.
Traceback (most recent call last):
File "...\Python36\lib\multiprocessing\process.py", line 258, in _bootstrap
self.run()
File "...\MSC\preprocess\parse_replay_info.py", line 44, in run
with self.run_config.start() as controller:
File "...\Python36\lib\site-packages\pysc2\run_configs\platforms.py", line 102, in start
version = version or FLAGS.sc2_version
File "...\Python36\lib\site-packages\absl\flags\_flagvalues.py", line 488, in __getattr__
raise _exceptions.UnparsedFlagAccessError(error_message)
absl.flags._exceptions.UnparsedFlagAccessError: Trying to access flag --sc2_version before flags were parsed.
As far as I can tell the error occurs because the flags from absl aren't passed to the child processes spawned to parse the replay info.
One temporary and tideous fix is to go into all the files used an hard code all the values instead of using the flag values. If someone else has a better fix, please let me know.
I don't know if it makes a difference, but my platform is Windows 10.
EDIT: I don't get the same error in Linux so the issue seems to be Windows specific.
Hi, thank you for implementing this repo!
Line 13 of the "parse_replay/sample_actions.py" is "from pysc2.lib import FUNCTIONS". However, such an import failed. I think it may be caused by DeepMind updated the pysc2.lib.
I appreciate it if you could check and fix this issue. Thanks
I have observed that the Google Drive link does not lead to a website that would be accessible by anyone.
Is this dataset no longer supported?
When I run the parse_replay_info.py, I get the following error:
pysc2.lib.remote_controller.RequestError: SC2APIProtocol.ResponseReplayInfo.Error.ParsingError: 'Could not open initData for the replay: C:\Users\DCS\AppData\Local\Temp\StarCraft II\TempReplayInfo~01.SC2Replay'
Could you please tell me how to fix this? Thank you.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.