Commits · 52c3083f10c0ec27183333765953f586bf55626b · yoogottamk / Flatland

May 16, 2019
- #26 pydeps · 52c3083f
  u214892 authored 5 years ago
  
  52c3083f
May 15, 2019
- Updated training: New state has two time frames · 8e79b68f
  Erik Nygren authored 5 years ago
  
  8e79b68f
- update requirements_dev.txt · 08c6d12c
  Erik Nygren authored 5 years ago
  
  08c6d12c
May 14, 2019
- added basic colors (from MPL colormap hsv) to SVG · 216d1ae6
  hagrid67 authored 5 years ago
  
  216d1ae6
- SVG trains bend around corners · 5bdf72e0
  hagrid67 authored 5 years ago
  
  5bdf72e0
- added SVG artwork in QTSVG renderer. · 1b3ca35a
  hagrid67 authored 5 years ago
  
  1b3ca35a
- Revert "trying to fix simulation infeaasibility" · 1dbb48be
  Guillaume Mollard authored 5 years ago
  
  This reverts commit 7be6f7c0.
  1dbb48be
- trying to fix simulation infeaasibility · 7be6f7c0
  Guillaume Mollard authored 5 years ago
  
  7be6f7c0
May 13, 2019
- latest training results · e163a42a
  Erik Nygren authored 5 years ago
  
  e163a42a
- fixed missing loop over agent when episode is solved · 0e139138
  Erik Nygren authored 5 years ago
  
  0e139138
- fixed formatting · 1169fcb6
  Erik Nygren authored 5 years ago
  
  1169fcb6
- fixed bug in reward function when all agents terminate · 36714e2c
  Erik Nygren authored 5 years ago
  
  36714e2c
- added split_tree function in order to simplify normalization of tree observation. · c2110f62
  Erik Nygren authored 5 years ago
  
  c2110f62
- removed debugging print in generator. Updated training: Agent that have... · a590c6b3
  Erik Nygren authored 5 years ago
  
  removed debugging print in generator. Updated training: Agent that have terminated do not receive observation untill all agents terminate.
  a590c6b3
May 12, 2019
- removed agents_handles and number_of_agents · f6d65e47
  hagrid67 authored 5 years ago
  
  Added load and save to env (left GridTransitionMap load and save for now)
  f6d65e47
- updated test_renderenv to use npz format. · b5b2ca8e
  hagrid67 authored 5 years ago
  
  Added a test for the PIL graphics layer.
  pre_agents_handles
  
  b5b2ca8e
- fixed up MPL GL again - colors in [0,1]^3, spurious plt.figure · c4848fb0
  hagrid67 authored 5 years ago
  
  c4848fb0
May 11, 2019
- improvements to editor2.ipynb · 026c3e10
  hagrid67 authored 5 years ago
  
  026c3e10
May 10, 2019
- Merge branch 'master' of gitlab.aicrowd.com:flatland/flatland · 302e0c76
  Erik Nygren authored 5 years ago
  
  302e0c76
- minor updates to training files · 58f7caf0
  Erik Nygren authored 5 years ago
  
  58f7caf0
May 09, 2019
- Merge branch 'env_manager' · d20778fc
  hagrid67 authored 5 years ago
  
  d20778fc
- moved agent_* lists to a list of EnvAgents · f2379642
  hagrid67 authored 5 years ago
  
  f2379642
- track *.pth files using Git LFS · e046035a
  hagrid67 authored 5 years ago
  
  e046035a
- add_agent_static · 6afc2a09
  hagrid67 authored 5 years ago
  
  6afc2a09
- add agent_utils · 95362985
  hagrid67 authored 5 years ago
  
  95362985
- complex rail gen: added extra connections · 71aea64e
  maljx authored 5 years ago
  
  71aea64e
- Add a link to the online documentation in the README · 417678af
  spmohanty authored 5 years ago
  
  417678af
- Use aws cp to copy files over · bb6ddb7a
  spmohanty authored 5 years ago
  
  bb6ddb7a
- Add a step to install flatland when building docs · 0d2a9486
  spmohanty authored 5 years ago
  
  0d2a9486
- Add a stage param to tests · e77e23ca
  spmohanty authored 5 years ago
  
  e77e23ca
- Addresses #2 : Setups CI to deploy docs to S3 · df2ba323
  spmohanty authored 5 years ago
  
  df2ba323
- Change copyright string to point to org instead of individual · 45efbdf1
  spmohanty authored 5 years ago
  
  45efbdf1
May 08, 2019
- bug fixes to complex rail gen · 2556ec15
  maljx authored 5 years ago
  
  2556ec15
- flake fixes · e42f4144
  maljx authored 5 years ago
  
  e42f4144
- updated complex generator to return start,goal,start dir · 47be1c5f
  maljx authored 5 years ago
  
  47be1c5f
- added Pillow to requirements · da35282a
  hagrid67 authored 5 years ago
  
  Added PIL graphics - 5-10 times faster than MPL move adaptColor into GraphicsLayer change scatter size arg to s like MPL move clf, get_cmap up into GraphicsLayer added boolean "train" arg to ddqn step() to make running faster in player / editor add plotTreeObs - not finished. improved pixel scaling in editor (I hope) Made the agent background steps update a progress bar.
  da35282a
- update frozen images used in tests · 1776a37b
  hagrid67 authored 5 years ago
  
  1776a37b
May 07, 2019
- random generator with turns · 8e3d61a9
  spiglerg authored 5 years ago
  
  8e3d61a9
- new generators that take agents_handles as extra input and return the agent's... · e2c3a494
  spiglerg authored 5 years ago
  
  new generators that take agents_handles as extra input and return the agent's initial position, direction and target
  e2c3a494
May 06, 2019
- added better agent network · f7501eb5
  Erik Nygren authored 5 years ago
  
  f7501eb5