Visualising AI state

[DorphBot]

The i-command is awesome, but work's extremely slow here. Even when i just add 2 of them, it takes 10-20 seconds after the last round for the program to terminate. Is that expected?

[akrewTik]

Hi there !

Thank you for this awesome job !

I'm quite a newbie on this and i can't have this tool work :/

I have replaced Ants.py and CanvasElement.js files but when i launch play_one_game. My commands do nothing nor in console nor in the visualizer.

In the html file i have something like : "overlay_history":[[],[[],[],[],[],

Does someone have an idea to help me please ?

P.S. : I have the starter java kit. I haven't done any work on this one.

Thank you very much !

[mac]

[Speed]

This is really awesome, thank you for your work

I tested the visualizer and had some trouble "installing" it. Is this way correct:
-extract the tools from the main page
-copy all files from github->aichallange/ants/visualizer and replace the ones from the tools
-copy and replace all .py files from github->aichallange/ants/
did I miss something?
edit: basicly same quastion as mac.

One problem however still remains: the lines I draw with line(row1,col1,row2,col2) dont use the shortest path (they dont wrap around the map). But you said this is corrected, so do I do something wrong, or is there an option for this?

edit says: how exatly is the syntax?
I output: "v line(10 10 60 60)" ist this correct?

has anyone wrapped the commands in c++ and does feel like sharing this?

[akrewTik]

[Tortortor]

[Speed]

[Speed]

[imargonis]

[jab_bott]

Hi all,

I've been busy this week so not had chance to respond...

'i' commands can get very slow, yes - this is because of the sheer amount of data involved (check the size of your replay html file). Each location on the map that you specify an 'i' command for adds an entry to a dictionary for each turn! So if you specify info for a mere 50x50 map locations over just 500 turns that's 1.25m entries (spread across the 500 dictionaries), each entry in the JSON will look something like "23x123":"info here", resulting in 25+ meg of JSON to be processed. The same applies to the 'v' commands really - combined the data can get very large and take a long time to process.

Regarding installation instructions - There's no need for a complicated 'installation' process with replacing files etc. - the way I do it is simply by cloning the entire repository from github and using that instead of the tools directory. I'll add an .md file to explain, as suggested by mac.