hiX SSKrSSKJrJr SSKrSSKrSSKJr SSK J r J r SSK J r Jr S\RS\4SjrS\\R$\R&4S\4S jrS"S\RS \S \SS4S jjrS#S \R.S \SS4SjjrS\R2S \RS\RSS4SjrS\R2SS4SjrS\R2S\4SjrS\S \R<S\SS4SjrS\\\\R@\!44S\R2S\"S\\\4SS4 Sjr#S\\$\\R@\!4S \RS\SS4Sjr%S#S \R.S \SS4Sjjr&S\R2S \RS\RSS4Sjr'S\R2SS4Sjr(S$S\R2S\SS4Sjjr)S%S\R2S\S \SS4S!jjr*g)&N)AnyUnion)spaces)check_for_nested_spacesis_image_space_channels_first) DummyVecEnv VecCheckNanspacereturnc`[U[R[R45(+$)zx Returns False if provided space is not representable as a single numpy array (e.g. Dict and Tuple spaces return False) ) isinstancerDictTupler s ^/home/james-whalen/.local/lib/python3.13/site-packages/stable_baselines3/common/env_checker.py_is_numpy_array_spacer s! %&++v||!<= ==c[R"UR[R"UR55$)z? Return False if a (Multi)Discrete space has a non-zero start. )npallclosestart zeros_likers r_starts_at_zerors& ;;u{{BMM%++$> ??r space_typekeyc ([U[R[R45(ac[ U5(dRU(aSUS3OSn[ R "[U5RSUSUSURSUS3 5 g g g ) a4 :param space: Observation or action space :param space_type: information about whether it is an observation or action space (for the warning message) :param key: When the observation space comes from a Dict space, we pass the corresponding key to have more precise warning messages. Defaults to "". z(key='z') z space z with a non-zero start (start=z) is not supported by Stable-Baselines3. You can use a wrapper (see https://stable-baselines3.readthedocs.io/en/master/guide/custom_env.html) or update your z space.N) r rDiscrete MultiDiscreterwarningswarntype__name__r)r rr maybe_keys r_check_non_zero_startr&s%&//6+?+?@AA/Z_J`J`(+fSE$  E{##$Aj\ Cabgbmbmano)\ 2 KaArobservation_spacecUR[R:wa'[R"SUSURS35 [R "UR S:g5(d([R "URS:g5(a[R"SUS35 Sn[U5(aSnURUS :dURS S :a[R"S 5 g g ) a2 Check that the input will be compatible with Stable-Baselines when the observation is apparently an image. :param observation_space: Observation space :param key: When the observation space comes from a Dict space, we pass the corresponding key to have more precise warning messages. Defaults to "". zIt seems that your observation z! is an image but its `dtype` is (z) whereas it has to be `np.uint8`. If your observation is not an image, we recommend you to flatten the observation to have only a 1D vectorrz%It seems that your observation space z is an image but the upper and lower bounds are not in [0, 255]. Because the CNN policy normalize automatically the observation you may encounter issue if the values are not in that range.$zThe minimal resolution for an image is 36x36 for the default `CnnPolicy`. You might need to use a custom features extractor cf. https://stable-baselines3.readthedocs.io/en/master/guide/custom_policy.htmlN) dtyperuint8r!r"anylowhighrshape)r'rnon_channel_idxs r_check_image_inputr4-s"((* -cU3$**+,' '  vv##q())RVV4E4J4Jc4Q-R-R 3C59K K O$%677/"48I8O8OPQ8RUW8W  ^ 9Xrenv action_spacec@[U[R5(apSnURR5H3upE[U[R5(aSn[ USU5 M5 U(a[ R "S5 [U[R5(aG[URR5S:a$[ R "SURS35 [U[R5(a[ R "S5 [ US5 [U[R5(a[ R "S 5 [ US 5 [U5(d[ R "S 5 g g ) zcEmit warnings when the observation space or action space used is not supported by Stable-Baselines.FT observationa8Nested observation spaces are not supported by Stable Baselines3 (Dict spaces inside Dict space). You should flatten it to have only one level of keys.For example, `dict(space1=dict(space2=Box(), space3=Box()), spaces4=Discrete())` is not supported but `dict(space2=Box(), spaces3=Box(), spaces4=Discrete())` is.r,zBThe MultiDiscrete observation space uses a multidimensional array z which is currently not supported by Stable-Baselines3. Please convert it to a 1D array using a wrapper: https://github.com/DLR-RM/stable-baselines3/issues/1836.zThe observation space is a Tuple, this is currently not supported by Stable Baselines3. However, you can convert it to a Dict observation space (cf. https://gymnasium.farama.org/api/spaces/composite/#dict). which is supported by SB3.zSequence observation space is not supported by Stable-Baselines3. You can pad your observation to have a fixed size instead. Note: The checks for returned values are skipped.actionzThe action space is not based off a numpy array. Typically this means it's either a Dict or Tuple space. This type of action space is currently not supported by Stable Baselines 3. You should try to flatten the action using a wrapper.N)r rritemsr&r!r"r lennvecr2rSequencer)r5r'r6 nested_dictrr s r_check_unsupported_spacesr?SsJ#V[[11 +2288:JC%--" !% <;  MMc #V%9%9::sCTCYCYC_C_?`cd?d PQbQgQgPhiG G #V\\22  ) +];#V__55  @ ,1  . .  &  /rc^[[U4Sj/55nUR5 [S5HHn[R "TR R5/5nURU5u nMJ g)z+Check for Inf and NaN using the VecWrapper.c>T$)Nr5sr_check_nan..ssr N) r rresetrangerarrayr6samplestep)r5vec_env_r9s` r _check_nanrNsa+{m45G MMO 2Y3++22456\\&) 1arc.[URS5$)zh Check if the env uses the convention for goal-conditioned envs (previously, the gym.GoalEnv interface) compute_reward)hasattr unwrappedrCs r _is_goal_envrSs 3=="2 33robs method_namec R[UR5S:dAS[UR5S[URR5535eSHGnX1R;dM[ SUSUS[URR5535e g) z Check that an environment implementing the `compute_rewards()` method (previously known as GoalEnv in gym) contains at least three elements, namely `observation`, `achieved_goal`, and `desired_goal`. zA goal conditioned env must contain at least 3 observation keys: `observation`, `achieved_goal`, and `desired_goal`. The current observation contains z keys: ) achieved_goal desired_goal!The observation returned by the `z3()` method of a goal-conditioned env requires the 'zA' key to be part of the observation dictionary. Current keys are N)r;rlistkeysAssertionError)rTr'rUrs r_check_goal_env_obsr^s  '' (A - ,,/0A0H0H,I+J'RVWhWoWoWtWtWvRwQx z - 1 .. . 3K=@stwsxy$$():)A)A)F)F)H$I#JL 1rrewardinfocrUSUSpTX!RXEU5:XdS5e[R"U5[R"U5pT[R"XD/5n[R"XU/5n[U[5(d[ UR 5S:Xa$URSS5nURSS5n[R"X3/5nURXgU5n U R S:XdSU R S 35eU SU:XdS U SS U35eg ) z` Check that reward is computed with `compute_reward` and that the implementation is vectorized. rXrYz3The reward was not computed with `compute_reward()`rr,)rbz7Unexpected shape for vectorized computation of reward: z != (2,)zBVectorized computation of reward differs from single computation:  != N)rPrrIr intr;r2reshape) rTr5r_r`rXrYbatch_achieved_goalsbatch_desired_goals batch_infosrewardss r_check_goal_env_compute_rewardrjs@#&o"6N8K< ''T =<= #%((="9288L;Q<88]$BC((L#?@-%%]-@-@)AQ)F3;;AqA199!Q?((D<(K  !5KXG ==D s$[\c\i\i[jjr"ss 1: ~#efmnofpeqquv|u}!~~ rc j[U[R5(d [U[5(a SUS35e[U[R5(a?[ R "[U5[ R5(d SUS35eO:[U5(a*[U[ R5(d SUS35e[U[ R5(GaURUR:Xd'SUSUSURSUR35e[ R"URUR5(d'SUS USURS UR35e[U[R5(aURUR pC[ R""[ R$"X:X:55nX:R'5(dX:R'5(anSUS US 3nU[)US 5S3- n[+U6H:nSR-[/[0U55nUSX7SUSXGSXS3 - nM< [3U5eUR5U5(d SUSU35eg)zX Check that the observation returned by the environment correspond to the declared one. rZz0()` method should be a single value, not a tuplezThe observation returned by `z()` method must be an intz ()` method must be a numpy arrayzC()` method does not match the shape of the given observation space z . Expected: z, actual shape: zU()` method does not match the data type (cannot cast) of the given observation space z, actual dtype: zD()` method does not match the bounds of the given observation space z. rz invalid indices: ,z Expected: z <= obs[z] <= z, actual value: z z6()` method does not match the given observation space N)r rrtuplerr issubdtyper#integerrndarrayr2can_castr-Boxr0r1where logical_orr/r;zipjoinmapstrr]contains) rTr'rU lower_bounds upper_boundsinvalid_indicesmessageindex index_strs r _check_obsrs ' 6 6    m .{m;k l m #V__55}}T#Y 33{7TU`Taaz5{{3 0 1 1#rzz**y.KK=Xx,yy*#rzz"" &&#))3 / }=..?-@A*0011A#)) N 3 {{399&7&=&=>> / }=..?-@A*0011A#)) N > ' 4 4):)>)>@Q@V@V, hhr}}S5GI['\]O"''))c.@-E-E-G-G7 }E66G5HNc/!"4566JKK /2E #S% 9I$\%8$9)ER^ReQfg)), C9G3%W--  % %c * * +K=966G5H J *rc[UR5S:Xa [X5 [UR5S;a[R"SUS35 gg)a Check that the observation space is correctly formatted when dealing with a ``Box()`` space. In particular, it checks: - that the dimensions are big enough when it is an image, and that the type matches - that the observation has an expected shape (warn the user if not) rW)r,rWzYour observation z has an unconventional shape (neither an image, nor a 1D vector). We recommend you to flatten the observation to have only a 1D vector or use a custom policy to properly process the data.N)r;r2r4r!r")r'rs r_check_box_obsrsY  " "#q(,2  " "#61 u%\ \ 2rcjUR5n[U[5(dS5e[U5S:XdS[U535eUupE[U[5(d SU35e[ U5(a.[U[ R5(de[XAS5 O[U[ R5(a[U[5(dS5eUR5UR R5:Xd7[SUR5SUR R535eUR R5H n[XFUR US5 M" O [XAS5 UR5nURU5n [U 5S :XdS [U 5S35eU upJpn[U[ R5(Ga[U[5(dS5e[ U5(aC[U[ R5(de[XAS5 [X@[!U 5U5 UR5UR R5:Xd7[SUR5SUR R535eUR R5H n[XFUR US5 M" O [XAS5 [U [ ["45(dS5e[U [$5(dS5e[U [$5(dS5e[U[5(dS5e[ U5(a2['US5(deXR)USUSU5:Xdeg g ![a!n[S US 3[U5-5UeS nAff=f![a!n[S US 3[U5-5UeS nAff=f)zT Check the returned values by the env when calling `.reset()` or `.step()` methods. z)`reset()` must return a tuple (obs, info)rbz9`reset()` must return a tuple of size 2 (obs, info), not zMThe second element of the tuple return by `reset()` must be a dictionary not rGz:The observation returned by `reset()` must be a dictionaryzRThe observation keys returned by `reset()` must match the observation space keys: rczError while checking key=z: Nz_The `step()` method must return five values: obs, reward, terminated, truncated, info. Actual: z values returned.z9The observation returned by `step()` must be a dictionaryrKzQThe observation keys returned by `step()` must match the observation space keys: z/The reward returned by `step()` must be a floatz)The `terminated` signal must be a booleanz(The `truncated` signal must be a booleanz;The `info` returned by `step()` must be a python dictionaryrPrXrY)rGr rmr;dictrSrrr^r\r]rrxrJrKrjfloatrdboolrQrP) r5r'r6 reset_returnsrTr`rer9datar_ terminated truncateds r_check_returned_valuesrs IIKM mU + +X-XX + }  "t&_`cdq`r_s$tt "IC dD ! !y%rswrx#yy !C+V[[9999CG< %v{{ 3 3#t$$b&bb$xxz.55::<< "xxzl$/@/G/G/L/L/N.OQ  %++002C Z38%6%=%=c%BGL3 373 "F 88F D t9> ==@YKGX Z> 04,C#V[[11#t$$a&aa$   /== ==  ? *3U6]D Ixxz.55::<< "xxzl$/@/G/G/L/L/N.OQ  %++002C Z38%6%=%=c%BFK3 362 fucl + +^-^^ + j$ ' 'T)TT ' i & &R(RR & dD ! !`#`` !Cs,----++C,@#nBUW[\\\\c" Z$'@R%H3q6%QRXYY ZH" Z$'@R%H3q6%QRXYY Zs0,O3P P#O??P P2P--P2cSn[US5(d SUS35e[US5(d SUS35e[UR[R5(d SUS35e[UR [R5(d SUS35e[ U5(a<[S 5 [UR[R5(dS 5eg g ) z Check that the observation and action spaces are defined and inherit from spaces.Space. For envs that follow the goal-conditioned standard (previously, the gym.GoalEnv interface) we check the observation space is gymnasium.spaces.Dict z,cf. https://gymnasium.farama.org/api/spaces/r'z'You must specify an observation space ()r6z"You must specify an action space (z:The observation space must inherit from gymnasium.spaces (z5The action space must inherit from gymnasium.spaces (zWe detected your env to be a GoalEnv because `env.compute_reward()` was defined. If it's not the case, please rename `env.compute_reward()` to something else to avoid False positives.zGoal conditioned envs (previously gym.GoalEnv) require the observation space to be gymnasium.spaces.Dict. Note: if your env is not a GoalEnv, please rename `env.compute_reward()` to something else to avoid False positive.N) rQr r'rSpacer6rSprintr)r5 gym_spacess r _check_spacesris @J 3+ , ,e0WXbWccd.ee , 3 ' '[+Mj\YZ)[[ '  v||  R CJa gf=f) a6 Check that an environment follows Gym API. This is particularly useful when using a custom environment. Please take a look at https://gymnasium.farama.org/api/env/ for more information about the API. It also optionally check that the environment is compatible with Stable-Baselines. :param env: The Gym environment that will be checked :param warn: Whether to output additional warnings mainly related to the interaction with Stable Baselines :param skip_render_check: Whether to skip the checks for the render method. True by default (useful for the CI) zdYour environment must inherit from the gymnasium.Env class cf. https://gymnasium.farama.org/api/env/r)seedz1The reset() method must accept a `seed` parameterNrr*r,zWe recommend you to use a symmetric and normalized Box action space (range=[-1, 1]) cf. https://stable-baselines3.readthedocs.io/en/master/guide/rl_tips.htmlz@Continuous action space must have a finite lower and upper boundzYour action space has dtype z5, we recommend using np.float32 to avoid cast errors.) r gymEnvrr'r6rG TypeErrorr?rrr:rrrrr/absr0r1r!r"allisfiniterIr-float32r=rrrrNNotImplementedError) r5r"rr'r6r obs_spacesrr s r check_envrs  SWW  nmn  #--##LT q   !#,G1;rsxiE>>$>@5&2F2F!FG@D@  3 UX bf $# &**# 3# # L4 3774 v||4 [a[g[g4 lp4 n*CGG**4cgg4$4TfkkPSX\( c5S)) *   sCx.   66E%rzz3676FLL6gj6os6r fjj s D (O]O]FLLO]X^XdXdO]imO]d sww 4 :swwdt2O 377O $O $O RVO r