Code

v1.0.0 alpha#154

Code

Draft

conorheins wants to merge 222 commits intomasterinfer-actively/pymdp:masterfrom v1.0.0_alphainfer-actively/pymdp:v1.0.0_alphaCopy head branch name to clipboard

+14,376-7,845

Commits

Commits on Jun 10, 2024

• added learning for online variational filtering

  dimarkov

  committed
  12bdc1d

• optimized learning routines to minimize calls to tree_map

  dimarkov

  committed
  754a336

• fix tests and passing of the results of the learning

  dimarkov

  committed
  eda7e01

• fix update state transition

  dimarkov

  committed
  147a37c

• move task.py to jax/envs/env.py

  

  Tim Verbelen

  committed
  bdf1e14

• fix names

  dimarkov

  committed
  71a3cb2

• notebook for showcasing inference and learning using different methods

  dimarkov

  committed
  65b33c8

• Merge pull request #133 from infer-actively/pymdp-move-PymdpEnv

  Show description for a82d7a1
  conorheins

  authored
  a82d7a1

• add a pyproject toml

  OzanCatalVerses

  committed
  19172db

• move task.py to jax/envs/env.py

  

  Tim Verbelen

  authored and

  conorheins

  committed
  948409a

• get a version that passes all tests

  OzanCatalVerses

  committed
  a1bc4b0

• Merge pull request #134 from infer-actively/feature/pyproject-toml

  Show description for be0a8f5
  conorheins

  authored
  be0a8f5

• added an example of A tensors and B tensors where smoothing should disambiguate hidden states in the past

  conorheins

  committed
  13ae015

• did first step of inference in the ovf example

  conorheins

  committed
  ae528d9

• add a pyproject toml

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  11a2f91

• get a version that passes all tests

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  f9f9bab

• add a pyproject toml

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  99f8253

• get a version that passes all tests

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  a975e7d

• initial graph worlds version

  

  Tim Verbelen

  committed
  9db1556

Commits on Jun 11, 2024

• named distribution implementation

  Show description for ad9c098
  alec-tschantz

  committed
  ad9c098

• updated example

  dimarkov

  committed
  6d83206

• Merge branch 'full-ovf-with-example' into full-ovf-implementation

  dimarkov

  committed
  b1e50d3

• removed oudated markdown cell at the end of the smoothing example notebook

  conorheins

  committed
  97b3716

• update graph worlds env

  

  Tim Verbelen

  committed
  980d86e

• add a first version of model compilation

  OzanCatalVerses

  committed
  0e0e1f5

• add missing s, add some basic tests

  OzanCatalVerses

  committed
  9f17816

• add unittest

  OzanCatalVerses

  committed
  630c59e

• rename compile to compile_model

  OzanCatalVerses

  committed
  3091da7

• added example notebook

  alec-tschantz

  committed
  465bfac

• add some documentation

  OzanCatalVerses

  committed
  4c07332

• api updates to jax/agent

  alec-tschantz

  committed
  a97f5d3

• Merge branch 'feature/distribution-api' of https://github.com/infer-actively/pymdp into feature/distribution-api

  Show description for f55a0a8
  alec-tschantz

  committed
  f55a0a8

• add get_dependencies

  OzanCatalVerses

  committed
  beca262

• clean up a bit

  OzanCatalVerses

  committed
  d070e37

• add dependency extraction to jax agent

  alec-tschantz

  committed
  3685a93

• acknowledge that the alphabet has an ordering

  OzanCatalVerses

  committed
  2fd13dc

• end to end test of agent API

  alec-tschantz

  committed
  f3bc163

• forgot to push notebook

  alec-tschantz

  committed
  49f10a6

• update random_B_matrix to handle complex action dependencies

  ran-weii

  committed
  5f40eb0

• change config naming to be more consistent

  OzanCatalVerses

  committed
  fb29fb3

• config naming changes

  OzanCatalVerses

  committed
  3381edc

• make the data argument optional

  Show description for 8aa2295
  OzanCatalVerses

  committed
  8aa2295

• updated agent api

  alec-tschantz

  committed
  eab270d

• updated notebook

  alec-tschantz

  committed
  bc591ae

• updated init for agent

  alec-tschantz

  committed
  df499f4

• add utils to convert combinations to flattened index with test

  ran-weii

  committed
  81ffea7

• updated example notebook

  alec-tschantz

  committed
  d7500d2

• added complicated example

  alec-tschantz

  committed
  b3beca5

• Merge branch 'feature/distribution-api' into feature/complex_act_dependency

  ran-weii

  committed
  7927648

• add rollout function and fix graph world env and demo notebook

  

  Tim Verbelen

  committed
  a9b4404

• initial solution for flattening B action dims

  ran-weii

  committed
  cf0058d

• move a copy of combination indexing utils to jax.utils

  ran-weii

  committed
  910dd89

• initial solution for multi action decoding

  ran-weii

  committed
  614115d

• vectorize combination indexing and update tests

  ran-weii

  committed
  af08d88

• handle no action dependency in B tensor flattening, use vectorized multi action encoding and decoding

  ran-weii

  committed
  179317e

• commit complex_action_dependency notebook

  ran-weii

  committed
  616fa59

Commits on Jun 12, 2024

• add knapsack_demo notebook, planning not working

  ran-weii

  committed
  58ef4af

• add function to construct flattend policies

  ran-weii

  committed
  21e3b58

• commit working knapsack demo notebook

  ran-weii

  committed
  bb1ac49

• add some docstring

  

  Tim Verbelen

  committed
  4626c74

• add some more comments in the notebook

  

  Tim Verbelen

  committed
  55144f9

• smoothing demo notebook starting to work with sparse BCOO arrays from jax.experimental.sparse

  conorheins

  committed
  0856bce

• in pymdp/jax/inference.py, added version of joint_dist_factor (sub-routine of smoothing_ovf) that dispatches on jnp.ndarrays and JAXSparse arrays (e.g., BCOO arrays)

  conorheins

  committed
  ccfe3d4

• add multimethod to the dependencies

  OzanCatalVerses

  committed
  84cdbf1

• Merge pull request #135 from infer-actively/feature/graph_env

  Show description for c11d2d2
  conorheins

  authored
  c11d2d2

• named distribution implementation

  Show description for 400fba2
  

  alec-tschantz

  authored and

  conorheins

  committed
  400fba2

• add a first version of model compilation

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  2692973

• add missing s, add some basic tests

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  ce89a0e

• add unittest

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  a742691

• rename compile to compile_model

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  98bd8c2

• added example notebook

  

  alec-tschantz

  authored and

  conorheins

  committed
  e930eb7

• api updates to jax/agent

  

  alec-tschantz

  authored and

  conorheins

  committed
  55f3141

• add some documentation

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  435dcb1

• add get_dependencies

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  91b1de7

• clean up a bit

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  a2bfbda

• add dependency extraction to jax agent

  

  alec-tschantz

  authored and

  conorheins

  committed
  ec97a13

• acknowledge that the alphabet has an ordering

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  0c63c36

• end to end test of agent API

  

  alec-tschantz

  authored and

  conorheins

  committed
  296795d

• forgot to push notebook

  

  alec-tschantz

  authored and

  conorheins

  committed
  ae878dc

• change config naming to be more consistent

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  45ae9a0

• config naming changes

  

  OzanCatalVerses

  authored and

  conorheins

  committed
  0a3194b

• make the data argument optional

  Show description for b838199
  

  OzanCatalVerses

  authored and

  conorheins

  committed
  b838199

• updated agent api

  

  alec-tschantz

  authored and

  conorheins

  committed
  6b5e8cc

• updated notebook

  

  alec-tschantz

  authored and

  conorheins

  committed
  ba55a07

• updated init for agent

  

  alec-tschantz

  authored and

  conorheins

  committed
  7f54738

• updated example notebook

  

  alec-tschantz

  authored and

  conorheins

  committed
  ac8f198

• added complicated example

  

  alec-tschantz

  authored and

  conorheins

  committed
  ff6a18f

• Merge branch 'feature/distribution-api' of https://github.com/infer-actively/pymdp into feature/distribution-api

  Show description for 98106a4
  alec-tschantz

  committed
  98106a4

• updated so we do not add a batch by default in agent constructor

  alec-tschantz

  committed
  1d26079

• fixed normalize

  alec-tschantz

  committed
  fac44bd

• function to rol out using jax

  toonvdm

  committed
  b10e328

• agent action perception loop

  toonvdm

  committed
  7524294

• Working state multiple cues

  toonvdm

  committed
  592169b

• Updated notebook

  toonvdm

  committed
  7e809e3

• made environment functional and added docstrings

  toonvdm

  committed
  3d5387b

• intermediate state

  toonvdm

  committed
  8bd70d8

• bug resolved demo

  

  salvatomm

  authored and

  toonvdm

  committed
  f96b00b

• Clean up notebook for the generalized tmaze

  toonvdm

  committed
  751f918

• remove duplicate line

  toonvdm

  committed
  9fbcae3

• remove duplicate

  toonvdm

  committed
  30233a9

• line length agent

  toonvdm

  committed
  1352a6b

• undo linting issues with agent.py

  toonvdm

  committed
  0ef6a39

• undo linting issues with agent.py

  toonvdm

  committed
  4ec0363

• smoothing example notebook now working without batch dimensions

  conorheins

  committed
  89208bb

• functional sparse array version of smoothing_ovf in inference.py (doesn't work with vmap!)

  conorheins

  committed
  9fbab82

• rename B_dependencies back to B_factor_list to not break tests

  ran-weii

  committed
  88a7aab

• fixed num_states / num_obs constructors

  alec-tschantz

  committed
  384548b

• update multi action encoding and decoing with test

  ran-weii

  committed
  1ec7123

• merge with distribution_api branch and resolve conflict

  ran-weii

  committed
  cb7ac8f

Commits on Jun 13, 2024

• Use new rollout signature

  toonvdm

  committed
  25ec151

• Merge pull request #137 from toonvdm/feature/agent-demo

  Show description for 727a75a
  dimarkov

  authored
  727a75a

• Merge remote-tracking branch 'origin/pymdp-verses-hackathon' into feature/distribution-api

  conorheins

  committed
  72c8510

• Merge pull request #136 from infer-actively/feature/distribution-api

  Show description for 01db186
  conorheins

  authored
  01db186

• Merge remote-tracking branch 'origin/pymdp-verses-hackathon' into HEAD

  conorheins

  committed
  4ff50a5

• Merge pull request #143 from infer-actively/feature/complex_act_dependency

  Show description for 07f37da
  conorheins

  authored
  07f37da

• Merge remote-tracking branch 'origin/pymdp-verses-hackathon' into feature/sparse_tensors_backend

  conorheins

  committed
  f30bf28

• don't overwrite C, D, E tensors if apply batch is True

  toonvdm

  committed
  449851e

• updated poetry.lock

  

  toonvdm

  authored and

  conorheins

  committed
  66c9301

• beginning of unit test file for sparse tensors backend

  conorheins

  committed
  c9493a2

• add sparse spm dot initial version

  OzanCatalVerses

  committed
  54374ae

• fixes to agent.py in pymdp.jax to deal with:

  Show description for 42e46f3
  conorheins

  committed
  42e46f3

• use apply_batch=False when constructing agent in generalized t maze demo notebook

  conorheins

  committed
  01165b1

• add a separate py file for tree searching

  

  Tim Verbelen

  committed
  c31bcec

• intermediate sharing some tree building

  

  Tim Verbelen

  committed
  5a5800b

• update

  

  Tim Verbelen

  committed
  031ce57

• step function

  

  alec-tschantz

  authored and

  Tim Verbelen

  committed
  f9c11a4

• vmap over policies

  

  alec-tschantz

  authored and

  Tim Verbelen

  committed
  e8b63cf

• integrated alec's new step function

  

  Tim Verbelen

  committed
  9b94a39

• we have a tree

  

  Tim Verbelen

  committed
  8c46cd8

• added vmap over leaves

  

  alec-tschantz

  authored and

  Tim Verbelen

  committed
  6cbfd06

• merging sophisticated inference backprop

  

  alec-tschantz

  authored and

  Tim Verbelen

  committed
  247d711

• vmap over observations

  

  alec-tschantz

  authored and

  Tim Verbelen

  committed
  a373439

• Merge pull request #144 from infer-actively/fix/error-c-tensor

  Show description for 639bf48
  conorheins

  authored
  639bf48

• fix spm_dot_sparse dims dims modifications

  OzanCatalVerses

  committed
  cb49577

• add sophisticated tmaze example

  

  Tim Verbelen

  committed
  b1b74aa

Commits on Jun 14, 2024

• Unit test for ovf smoothing using sparse tensors

  toonvdm

  committed
  59131c6

• updated docstring for todense hack

  toonvdm

  committed
  0a3faae

• get the inference methods comparison notebook to work with latest hackathon version

  OzanCatalVerses

  committed
  a0967dd

• benchmarking notebook

  toonvdm

  committed
  2edabf6

Commits on Jun 17, 2024

• fix multimethod matching

  OzanCatalVerses

  committed
  37b253d

Commits on Jun 18, 2024

• change array->arraylike in all type signatures

  OzanCatalVerses

  committed
  787fb00

Commits on Jun 21, 2024

• unit tests for learning + B_dependencies in update_state_transition_dirichlet

  toonvdm

  committed
  d21d8b0

Commits on Jun 22, 2024

• added learning from main + added docstring + unit tested jax implementation against numpy counterpart

  toonvdm

  committed
  b7bbbf9

• selective update of factors for learning the transition parameters

  toonvdm

  committed
  6b7d50f

Commits on Jun 23, 2024

• return all transition matrices when factors_to_update is not all

  toonvdm

  committed
  03b2cd1

• return all transition matrices when factors_to_update is not all

  toonvdm

  committed
  568900d

• remove legacy comment

  toonvdm

  committed
  02cd58c

Commits on Jun 27, 2024

• Merge pull request #148 from infer-actively/feature/learning-unit-tests

  Show description for 256f143
  dimarkov

  authored
  256f143

• add imageio dependency

  

  Tim Verbelen

  committed
  2a1c34b

• also allow to use Distribution for C and D

  

  Tim Verbelen

  committed
  8d3bc80

• Added Leaf class to register arbitrary objects as leaf

  toonvdm

  committed
  2ad121f

• comment policies multi

  toonvdm

  committed
  92b5e4d

• Resolve merge conflicts

  toonvdm

  committed
  0cfc999

• add demo notebooks for the two jax envs

  

  Tim Verbelen

  committed
  bee8872

• Merge branch 'feature/sophisticated-jax-rebased' into hackaton-demos

  

  Tim Verbelen

  committed
  3ba3143

• refactor si code and examples

  

  Tim Verbelen

  committed
  d1b5a0e

• move generate graph function to env file

  

  Tim Verbelen

  committed
  c249dd5

• running sophisticated inference in a generalized tmaze env

  

  dimarkov

  authored and

  Tim Verbelen

  committed
  2605d36

• add mctx based mcts planner

  

  Tim Verbelen

  committed
  11a3e4d

• simpler graph world demo

  

  Tim Verbelen

  committed
  84c9943

• use custom scan to use scan signature to iterate over sparse arrays

  toonvdm

  committed
  086b881

• Merge branch 'feature/sparse_tensors_backend' into hackaton-demos

  

  Tim Verbelen

  committed
  1b129bc

• create separate subdir for the sparse experiment

  

  Tim Verbelen

  committed
  1273241

Commits on Jun 28, 2024

• Added mctx and mediapy dependencies

  toonvdm

  committed
  092beed

• generated outputs for benchmark notebook

  toonvdm

  committed
  927767e

Commits on Jul 31, 2024

• add readable string repr for Distribution

  tverbele

  committed
  4da4476

• add Model class for describing the full model in api

  tverbele

  committed
  bac99ed

Commits on Aug 2, 2024

• fix bug in creating preference distribution

  tverbele

  committed
  f91c1ab

Commits on Aug 8, 2024

• also allow specifying state preferences (H) through model/distribution

  tverbele

  committed
  4dfe790

• allow to set gamma and make default value 1

  tverbele

  committed
  d7f605a

Commits on Aug 9, 2024

• use muzero_policy as default instead of the gumbel_muzero_policy

  tverbele

  committed
  8bf9a38

Commits on Sep 20, 2024

• Merge pull request #149 from infer-actively/api_model

  Show description for 3f1287a
  conorheins

  authored
  3f1287a

• Merge pull request #151 from infer-actively/si_fixes

  Show description for 75123df
  conorheins

  authored
  75123df

• Merged @dimarkov's upstream changes to master into hackaton_demos

  Show description for 6a46032
  

  conorheins

  and

  tverbele

  committed
  6a46032

• Merge pull request #152 from infer-actively/hackaton-demos

  Show description for 707dd43
  conorheins

  authored
  707dd43

• fix Distribution API notebook

  

  Tim Verbelen

  committed
  dfdd9d0

• fixed the initialization of the E vector in Agent constructor so that it cannot be a list, must be an array or Distribution instance

  conorheins

  committed
  908aa8a

• Added apply_batch=False flag to all initializations of an Agent in the inference_methods_comparison.ipynb notebook

  conorheins

  committed
  368907f

• fix Graph Worlds demo

  

  Tim Verbelen

  committed
  cb78e18

• changes to control.py:

  Show description for c9031f9
  conorheins

  committed
  c9031f9

• fixed deprecated syntax of Agent.infer_states in mcts.py and si.py

  conorheins

  committed
  5a09229

• fixes to sparse_benchmark.ipynb notebook:

  Show description for c4e05ff
  conorheins

  committed
  c4e05ff

Commits on Sep 23, 2024

• cleanup the dependencies and poetry .toml

  

  Tim Verbelen

  committed
  18aac74

• remove unused pandas import

  

  Tim Verbelen

  committed
  249db72

• fix distribution unit test

  

  Tim Verbelen

  committed
  a741886

• fix update_empirical_prior and infer_states in tests

  

  Tim Verbelen

  committed
  d9c9a83

• configure setuptools using pyproject.toml

  

  Tim Verbelen

  committed
  ff6e27a

• fix typo

  

  Tim Verbelen

  committed
  d41fccc

Commits on Sep 24, 2024

• declare packages

  tverbele

  committed
  e0d63c3

• remove requirements.txt from README

  tverbele

  committed
  f9049da

• added back in decode_multi_actions as method of Agent, originally from @ran-weii's complex_act_dependencies branch

  conorheins

  committed
  5afbf38

• added optional factors_to_update argument into pymdp.jax.learning.update_state_likelihood_dirichlet() function

  conorheins

  committed
  c9630af

• fixed unit test test_sparse_smoothing in test_jax_sparse_backend to respect new output structure of pymdp.jax.inference.smoothing_ovf

  conorheins

  committed
  6652e39

• removed calls to np.random.seed(0) in test_learning_jax.py

  conorheins

  committed
  89ac754

• added more authors to author list in pyproject.toml (unfinished/WIP)

  conorheins

  committed
  58590de

• fixed update_state_likelihood_dirichlet to allow for factors_to_update to work properly, not using tree_map anymore

  conorheins

  committed
  265e50a

• Refactor of pymdp so that jax is the main backend, and numpy version now relegated to a legacy folder.

  Show description for 57d730d
  

  conorheins

  and

  tverbele

  committed
  57d730d

• fix refactored packages

  

  Tim Verbelen

  committed
  cc48366

• removed field(static=True) option for policies matrix in Agent

  conorheins

  committed
  d235ca6

• reimplementation of missing line when checking pA in self.validate() method of Agent, from commit 1958e4d

  conorheins

  committed
  b882927

• update to pymdp.Agent:

  Show description for fd44804
  conorheins

  committed
  fd44804

Commits on Sep 30, 2024

• changed github workflow configuration .yml to no longer test on Python 3.10 and to not assume a requirements.txt file. Authored by @tverbele

  conorheins

  committed
  03991da

Commits on Oct 2, 2024

• - renamed PyMDPEnv to Env throughout the repo

  Show description for da3716c
  

  conorheins

  and

  tverbele

  committed
  da3716c

• renamed PyMDPEnv to Env in learning_gridworld.ipynb notebook

  conorheins

  committed
  2d9e755

• renamed PyMDPEnv to Env in graph_worlds.py

  conorheins

  committed
  c3b795b

• renamed PyMDPEnv to Env in grid_world_demo.ipynb

  conorheins

  committed
  786ec50

• started refactoring GeneralizedTMaze(Env) class

  conorheins

  committed
  b4513b4

• removed .jax module prefixes from imports in pymdp.envs.rollout

  conorheins

  committed
  99cf06e

• refactor legacy TMaze to jax Env

  

  Tim Verbelen

  committed
  591bf24

• add TMaze demo notebook

  

  Tim Verbelen

  committed
  696ffe9

• refactor PyMDPEnv to Env

  

  Tim Verbelen

  committed
  9733895

• Merge branch 'tmaze' into v1.0.0_alpha

  

  Tim Verbelen

  committed
  18f6903

• fix reset

  

  Tim Verbelen

  committed
  a958270

Commits on Oct 3, 2024

• add pymdp.envs.assets to package

  

  Tim Verbelen

  committed
  82b80e5

• package image files in assets

  

  Tim Verbelen

  committed
  0e644f1

• fix state initialization on reset

  

  Tim Verbelen

  committed
  d081c23

Commits on Nov 18, 2024

• flatten pB action dims for parameter learning with multi actions

  ran-wei-verses

  committed
  6a0c2b0

Commits on Dec 2, 2024

• Merge pull request #161 from ran-wei-verses/v1.0.0_alpha

  Show description for fa5213a
  conorheins

  authored
  fa5213a

Commits on Dec 7, 2024

• Set minimum Python version to 3.10 for improved compatibility and feature support

  nikolamilovic1

  committed
  b49587d

• Add setup.cfg

  nikolamilovic1

  committed
  66ae9f3

• Modify setup.cfg to be equivalent to pyproject.toml.

  nikolamilovic1

  committed
  af28a0f

Commits on Dec 18, 2024

• Merge pull request #162 from nikolamilovic-ft/v1.0.0_alpha

  Show description for 3f06266
  conorheins

  authored
  3f06266