diff --git a/api/_modules/aepsych/config.html b/api/_modules/aepsych/config.html index c6f99946c..ba568c868 100644 --- a/api/_modules/aepsych/config.html +++ b/api/_modules/aepsych/config.html @@ -200,8 +200,8 @@

Source code for aepsych.config

                     # Validate the parameter-specific block
                     self._check_param_settings(par_name)
 
-                    lb[i] = self[par_name]["lower_bound"]
-                    ub[i] = self[par_name]["upper_bound"]
+                    lb[i] = self[par_name].get("lower_bound", fallback="0")
+                    ub[i] = self[par_name].get("upper_bound", fallback="1")
 
                 self["common"]["lb"] = f"[{', '.join(lb)}]"
                 self["common"]["ub"] = f"[{', '.join(ub)}]"
@@ -278,8 +278,45 @@ 
Source code for aepsych.config
                 raise ValueError(
                     f"Parameter {param_name} is missing the upper_bound setting."
                 )
+
+        elif param_block["par_type"] == "integer":
+            # Check if bounds exist and actaully integers
+            if "lower_bound" not in param_block:
+                raise ValueError(
+                    f"Parameter {param_name} is missing the lower_bound setting."
+                )
+            if "upper_bound" not in param_block:
+                raise ValueError(
+                    f"Parameter {param_name} is missing the upper_bound setting."
+                )
+
+            try:
+                if not (
+                    self.getint(param_name, "lower_bound") % 1 == 0
+                    and self.getint(param_name, "upper_bound") % 1 == 0
+                ):
+                    raise ParameterConfigError(
+                        f"Parameter {param_name} has non-integer bounds."
+                    )
+            except ValueError:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} has non-discrete bounds."
+                )
+
+        elif param_block["par_type"] == "binary":
+            if "lower_bound" in param_block or "upper_bound" in param_block:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} is binary and shouldn't have bounds."
+                )
+
+        elif param_block["par_type"] == "fixed":
+            if "value" not in param_block:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} is fixed and needs to have value set."
+                )
+
         else:
-            raise ValueError(
+            raise ParameterConfigError(
                 f"Parameter {param_name} has an unsupported parameter type {param_block['par_type']}."
             )
 
@@ -510,6 +547,10 @@ 
Source code for aepsych.config
 Config.register_module(botorch.acquisition)
 Config.register_module(botorch.acquisition.multi_objective)
 Config.registered_names["None"] = None
+
+
+
[docs]class ParameterConfigError(Exception):
+    pass
diff --git a/api/_modules/aepsych/config/index.html b/api/_modules/aepsych/config/index.html index c6f99946c..ba568c868 100644 --- a/api/_modules/aepsych/config/index.html +++ b/api/_modules/aepsych/config/index.html @@ -200,8 +200,8 @@

Source code for aepsych.config

                     # Validate the parameter-specific block
                     self._check_param_settings(par_name)
 
-                    lb[i] = self[par_name]["lower_bound"]
-                    ub[i] = self[par_name]["upper_bound"]
+                    lb[i] = self[par_name].get("lower_bound", fallback="0")
+                    ub[i] = self[par_name].get("upper_bound", fallback="1")
 
                 self["common"]["lb"] = f"[{', '.join(lb)}]"
                 self["common"]["ub"] = f"[{', '.join(ub)}]"
@@ -278,8 +278,45 @@ 
Source code for aepsych.config
                 raise ValueError(
                     f"Parameter {param_name} is missing the upper_bound setting."
                 )
+
+        elif param_block["par_type"] == "integer":
+            # Check if bounds exist and actaully integers
+            if "lower_bound" not in param_block:
+                raise ValueError(
+                    f"Parameter {param_name} is missing the lower_bound setting."
+                )
+            if "upper_bound" not in param_block:
+                raise ValueError(
+                    f"Parameter {param_name} is missing the upper_bound setting."
+                )
+
+            try:
+                if not (
+                    self.getint(param_name, "lower_bound") % 1 == 0
+                    and self.getint(param_name, "upper_bound") % 1 == 0
+                ):
+                    raise ParameterConfigError(
+                        f"Parameter {param_name} has non-integer bounds."
+                    )
+            except ValueError:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} has non-discrete bounds."
+                )
+
+        elif param_block["par_type"] == "binary":
+            if "lower_bound" in param_block or "upper_bound" in param_block:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} is binary and shouldn't have bounds."
+                )
+
+        elif param_block["par_type"] == "fixed":
+            if "value" not in param_block:
+                raise ParameterConfigError(
+                    f"Parameter {param_name} is fixed and needs to have value set."
+                )
+
         else:
-            raise ValueError(
+            raise ParameterConfigError(
                 f"Parameter {param_name} has an unsupported parameter type {param_block['par_type']}."
             )
 
@@ -510,6 +547,10 @@ 
Source code for aepsych.config
 Config.register_module(botorch.acquisition)
 Config.register_module(botorch.acquisition.multi_objective)
 Config.registered_names["None"] = None
+
+
+
[docs]class ParameterConfigError(Exception):
+    pass
diff --git a/api/_modules/aepsych/generators/manual_generator.html b/api/_modules/aepsych/generators/manual_generator.html index 5cffef88c..9268be4f9 100644 --- a/api/_modules/aepsych/generators/manual_generator.html +++ b/api/_modules/aepsych/generators/manual_generator.html @@ -24,9 +24,8 @@

Source code for aepsych.generators.manual_generator

# LICENSE file in the root directory of this source tree. import warnings -from typing import Dict, Optional, Union +from typing import Dict, Optional -import numpy as np import torch from aepsych.config import Config from aepsych.generators.base import AEPsychGenerator @@ -44,7 +43,7 @@

Source code for aepsych.generators.manual_generator

self, lb: torch.Tensor, ub: torch.Tensor, - points: Union[np.ndarray, torch.Tensor], + points: torch.Tensor, dim: Optional[int] = None, shuffle: bool = True, seed: Optional[int] = None, @@ -53,16 +52,18 @@

Source code for aepsych.generators.manual_generator

Args: lb torch.Tensor: Lower bounds of each parameter. ub torch.Tensor: Upper bounds of each parameter. - points (Union[np.ndarray, torch.Tensor]): The points that will be generated. + points torch.Tensor: The points that will be generated. dim (int, optional): Dimensionality of the parameter space. If None, it is inferred from lb and ub. shuffle (bool): Whether or not to shuffle the order of the points. True by default. """ self.seed = seed self.lb, self.ub, self.dim = _process_bounds(lb, ub, dim) + self.points = points if shuffle: - np.random.seed(self.seed) - np.random.shuffle(points) - self.points = torch.tensor(points) + if seed is not None: + torch.manual_seed(seed) + self.points = points[torch.randperm(len(points))] + self.max_asks = len(self.points) self._idx = 0 @@ -100,10 +101,14 @@

Source code for aepsych.generators.manual_generator

lb = config.gettensor(name, "lb") ub = config.gettensor(name, "ub") dim = config.getint(name, "dim", fallback=None) - points = config.getarray(name, "points") + points = config.gettensor(name, "points") shuffle = config.getboolean(name, "shuffle", fallback=True) seed = config.getint(name, "seed", fallback=None) + if len(points.shape) == 3: + # Configs have a reasonable natural input method that produces incorrect tensors + points = points.swapaxes(-1, -2) + options = { "lb": lb, "ub": ub, @@ -125,10 +130,10 @@

Source code for aepsych.generators.manual_generator

def __init__( self, - lb: Union[np.ndarray, torch.Tensor], - ub: Union[np.ndarray, torch.Tensor], - window: Union[np.ndarray, torch.Tensor], - points: Union[np.ndarray, torch.Tensor], + lb: torch.Tensor, + ub: torch.Tensor, + window: torch.Tensor, + points: torch.Tensor, samples_per_point: int, dim: Optional[int] = None, shuffle: bool = True, @@ -138,7 +143,10 @@

Source code for aepsych.generators.manual_generator

Args: lb (Union[np.ndarray, torch.Tensor]): Lower bounds of each parameter. ub (Union[np.ndarray, torch.Tensor]): Upper bounds of each parameter. - window (Union[np.ndarray, torch.Tensor]): How far away to sample from the reference point along each dimension. + window (Union[np.ndarray, torch.Tensor]): How far away to sample from the + reference point along each dimension. If the parameters are transformed, + the proportion of the range (based on ub/lb given) covered by the window + will be preserved (and not the absolute distance from the reference points). points (Union[np.ndarray, torch.Tensor]): The points that will be generated. samples_per_point (int): How many samples around each point to take. dim (int, optional): Dimensionality of the parameter space. If None, it is inferred from lb and ub. @@ -146,16 +154,27 @@

Source code for aepsych.generators.manual_generator

seed (int, optional): Random seed. """ lb, ub, dim = _process_bounds(lb, ub, dim) - points = torch.Tensor(points) self.engine = SobolEngine(dimension=dim, scramble=True, seed=seed) - generated = [] + gen_points = [] + if len(points.shape) > 2: + # We need to determine how many stimuli there are per trial to maintain the proper tensor shape + n_draws = points.shape[-1] + else: + n_draws = 1 for point in points: + if len(points.shape) > 2: + point = point.T p_lb = torch.max(point - window, lb) p_ub = torch.min(point + window, ub) - grid = self.engine.draw(samples_per_point) - grid = p_lb + (p_ub - p_lb) * grid - generated.append(grid) - generated = torch.Tensor(np.vstack(generated)) # type: ignore + for _ in range(samples_per_point): + grid = self.engine.draw(n_draws) + grid = p_lb + (p_ub - p_lb) * grid + gen_points.append(grid) + if len(points.shape) > 2: + generated = torch.stack(gen_points) + generated = generated.swapaxes(-2, -1) + else: + generated = torch.vstack(gen_points) super().__init__(lb, ub, generated, dim, shuffle, seed) # type: ignore diff --git a/api/_modules/aepsych/generators/manual_generator/index.html b/api/_modules/aepsych/generators/manual_generator/index.html index 5cffef88c..9268be4f9 100644 --- a/api/_modules/aepsych/generators/manual_generator/index.html +++ b/api/_modules/aepsych/generators/manual_generator/index.html @@ -24,9 +24,8 @@

Source code for aepsych.generators.manual_generator

# LICENSE file in the root directory of this source tree. import warnings -from typing import Dict, Optional, Union +from typing import Dict, Optional -import numpy as np import torch from aepsych.config import Config from aepsych.generators.base import AEPsychGenerator @@ -44,7 +43,7 @@

Source code for aepsych.generators.manual_generator

self, lb: torch.Tensor, ub: torch.Tensor, - points: Union[np.ndarray, torch.Tensor], + points: torch.Tensor, dim: Optional[int] = None, shuffle: bool = True, seed: Optional[int] = None, @@ -53,16 +52,18 @@

Source code for aepsych.generators.manual_generator

Args: lb torch.Tensor: Lower bounds of each parameter. ub torch.Tensor: Upper bounds of each parameter. - points (Union[np.ndarray, torch.Tensor]): The points that will be generated. + points torch.Tensor: The points that will be generated. dim (int, optional): Dimensionality of the parameter space. If None, it is inferred from lb and ub. shuffle (bool): Whether or not to shuffle the order of the points. True by default. """ self.seed = seed self.lb, self.ub, self.dim = _process_bounds(lb, ub, dim) + self.points = points if shuffle: - np.random.seed(self.seed) - np.random.shuffle(points) - self.points = torch.tensor(points) + if seed is not None: + torch.manual_seed(seed) + self.points = points[torch.randperm(len(points))] + self.max_asks = len(self.points) self._idx = 0 @@ -100,10 +101,14 @@

Source code for aepsych.generators.manual_generator

lb = config.gettensor(name, "lb") ub = config.gettensor(name, "ub") dim = config.getint(name, "dim", fallback=None) - points = config.getarray(name, "points") + points = config.gettensor(name, "points") shuffle = config.getboolean(name, "shuffle", fallback=True) seed = config.getint(name, "seed", fallback=None) + if len(points.shape) == 3: + # Configs have a reasonable natural input method that produces incorrect tensors + points = points.swapaxes(-1, -2) + options = { "lb": lb, "ub": ub, @@ -125,10 +130,10 @@

Source code for aepsych.generators.manual_generator

def __init__( self, - lb: Union[np.ndarray, torch.Tensor], - ub: Union[np.ndarray, torch.Tensor], - window: Union[np.ndarray, torch.Tensor], - points: Union[np.ndarray, torch.Tensor], + lb: torch.Tensor, + ub: torch.Tensor, + window: torch.Tensor, + points: torch.Tensor, samples_per_point: int, dim: Optional[int] = None, shuffle: bool = True, @@ -138,7 +143,10 @@

Source code for aepsych.generators.manual_generator

Args: lb (Union[np.ndarray, torch.Tensor]): Lower bounds of each parameter. ub (Union[np.ndarray, torch.Tensor]): Upper bounds of each parameter. - window (Union[np.ndarray, torch.Tensor]): How far away to sample from the reference point along each dimension. + window (Union[np.ndarray, torch.Tensor]): How far away to sample from the + reference point along each dimension. If the parameters are transformed, + the proportion of the range (based on ub/lb given) covered by the window + will be preserved (and not the absolute distance from the reference points). points (Union[np.ndarray, torch.Tensor]): The points that will be generated. samples_per_point (int): How many samples around each point to take. dim (int, optional): Dimensionality of the parameter space. If None, it is inferred from lb and ub. @@ -146,16 +154,27 @@

Source code for aepsych.generators.manual_generator

seed (int, optional): Random seed. """ lb, ub, dim = _process_bounds(lb, ub, dim) - points = torch.Tensor(points) self.engine = SobolEngine(dimension=dim, scramble=True, seed=seed) - generated = [] + gen_points = [] + if len(points.shape) > 2: + # We need to determine how many stimuli there are per trial to maintain the proper tensor shape + n_draws = points.shape[-1] + else: + n_draws = 1 for point in points: + if len(points.shape) > 2: + point = point.T p_lb = torch.max(point - window, lb) p_ub = torch.min(point + window, ub) - grid = self.engine.draw(samples_per_point) - grid = p_lb + (p_ub - p_lb) * grid - generated.append(grid) - generated = torch.Tensor(np.vstack(generated)) # type: ignore + for _ in range(samples_per_point): + grid = self.engine.draw(n_draws) + grid = p_lb + (p_ub - p_lb) * grid + gen_points.append(grid) + if len(points.shape) > 2: + generated = torch.stack(gen_points) + generated = generated.swapaxes(-2, -1) + else: + generated = torch.vstack(gen_points) super().__init__(lb, ub, generated, dim, shuffle, seed) # type: ignore diff --git a/api/_modules/aepsych/strategy.html b/api/_modules/aepsych/strategy.html index c4e2d297a..7192f15ca 100644 --- a/api/_modules/aepsych/strategy.html +++ b/api/_modules/aepsych/strategy.html @@ -256,7 +256,7 @@

Source code for aepsych.strategy

         if self.stimuli_per_trial == 1:
             self.event_shape: Tuple[int, ...] = (self.dim,)
 
-        if self.stimuli_per_trial == 2:
+        if self.stimuli_per_trial > 1:
             self.event_shape = (self.dim, self.stimuli_per_trial)
 
         self.model = model
diff --git a/api/_modules/aepsych/strategy/index.html b/api/_modules/aepsych/strategy/index.html
index c4e2d297a..7192f15ca 100644
--- a/api/_modules/aepsych/strategy/index.html
+++ b/api/_modules/aepsych/strategy/index.html
@@ -256,7 +256,7 @@ 
Source code for aepsych.strategy
         if self.stimuli_per_trial == 1:
             self.event_shape: Tuple[int, ...] = (self.dim,)
 
-        if self.stimuli_per_trial == 2:
+        if self.stimuli_per_trial > 1:
             self.event_shape = (self.dim, self.stimuli_per_trial)
 
         self.model = model
diff --git a/api/_modules/aepsych/utils.html b/api/_modules/aepsych/utils.html
index 95f1d534c..403977237 100644
--- a/api/_modules/aepsych/utils.html
+++ b/api/_modules/aepsych/utils.html
@@ -25,7 +25,7 @@ 
Source code for aepsych.utils
 
 from collections.abc import Iterable
 from configparser import NoOptionError
-from typing import Any, Dict, List, Mapping, Optional, Tuple, Union
+from typing import Any, Dict, Mapping, Optional, Tuple, Union
 
 import numpy as np
 import torch
@@ -250,120 +250,35 @@ 
Source code for aepsych.utils
     return result

 
 
-def _get_ax_parameters(
-    config: Config,
-) -> Tuple[list[Dict[str, Any]], list[Dict[str, Any]], list[Dict[str, Any]]]:
-    range_parnames = config.getlist("common", "parnames", element_type=str, fallback=[])
-    lb = config.getlist("common", "lb", element_type=float, fallback=[])
-    ub = config.getlist("common", "ub", element_type=float, fallback=[])
-
-    assert (
-        len(range_parnames) == len(lb) == len(ub)
-    ), f"Length of parnames ({range_parnames}), lb ({lb}), and ub ({ub}) don't match!"
-
-    range_params = [
-        {
-            "name": parname,
-            "type": "range",
-            "value_type": config.get(str(parname), "value_type", fallback="float"),
-            "log_scale": config.getboolean(str(parname), "log_scale", fallback=False),
-            "bounds": [l, u],
-        }
-        for parname, l, u in zip(range_parnames, lb, ub)
-    ]
-
-    choice_parnames = config.getlist(
-        "common", "choice_parnames", element_type=str, fallback=[]
-    )
-    choices = [
-        config.getlist(
-            str(parname), "choices", element_type=str, fallback=["True", "False"]
+
[docs]def get_bounds(config: Config) -> torch.Tensor:
+    r"""Return the bounds for all parameters in config.
+
+    Args:
+        config (Config): The config to find the bounds from.
+
+    Returns:
+        torch.Tensor: A `[2, d]` tensor with the lower and upper bounds for each
+            parameter.
+    """
+    parnames = config.getlist("common", "parnames", element_type=str)
+
+    # Try to build a full array of bounds based on parameter-specific bounds
+    try:
+        _lower_bounds = torch.tensor(
+            [config.getfloat(par, "lower_bound") for par in parnames]
+        )
+        _upper_bounds = torch.tensor(
+            [config.getfloat(par, "upper_bound") for par in parnames]
         )
-        for parname in choice_parnames
-    ]
-    choice_params = [
-        {
-            "name": parname,
-            "type": "choice",
-            "value_type": config.get(str(parname), "value_type", fallback="str"),
-            "is_ordered": config.getboolean(str(parname), "is_ordered", fallback=False),
-            "values": choice,
-        }
-        for parname, choice in zip(choice_parnames, choices)
-    ]
-
-    fixed_parnames = config.getlist(
-        "common", "fixed_parnames", element_type=str, fallback=[]
-    )
-    values = []
-    for parname in fixed_parnames:
-        try:
-            try:
-                value: Union[float, str] = config.getfloat(str(parname), "value")
-            except ValueError:
-                value = config.get(str(parname), "value")
-
-            values.append(value)
-        except NoOptionError:
-            raise RuntimeError(f"Missing value for fixed parameter {parname}!")
-    fixed_params = [
-        {
-            "name": parname,
-            "type": "fixed",
-            "value": value,
-        }
-        for parname, value in zip(fixed_parnames, values)
-    ]
-
-    return range_params, choice_params, fixed_params
-
-
-
[docs]def get_parameters(config: Config) -> List[Dict]:
-    range_params, choice_params, fixed_params = _get_ax_parameters(config)
-    return range_params + choice_params + fixed_params

 
+        bounds = torch.stack((_lower_bounds, _upper_bounds))
 
-
[docs]def get_bounds(config: Config) -> torch.Tensor:
-    range_params, choice_params, _ = _get_ax_parameters(config)
-    # Need to sum dimensions added by both range and choice parameters
-    bounds = [parm["bounds"] for parm in range_params]
-    for par in choice_params:
-        n_vals = len(par["values"])
-        if par["is_ordered"]:
-            bounds.append(
-                [0, 1]
-            )  # Ordered choice params are encoded like continuous parameters
-        elif n_vals > 2:
-            for _ in range(n_vals):
-                bounds.append(
-                    [0, 1]
-                )  # Choice parameter is one-hot encoded such that they add 1 dim for every choice
-        else:
-            for _ in range(n_vals - 1):
-                bounds.append(
-                    [0, 1]
-                )  # Choice parameters with n_choices <= 2 add n_choices - 1 dims
-
-    return torch.tensor(bounds)

-
-
-
[docs]def get_dim(config: Config) -> int:
-    range_params, choice_params, _ = _get_ax_parameters(config)
-    # Need to sum dimensions added by both range and choice parameters
-    dim = len(range_params)  # 1 dim per range parameter
-    for par in choice_params:
-        if par["is_ordered"]:
-            dim += 1  # Ordered choice params are encoded like continuous parameters
-        elif len(par["values"]) > 2:
-            dim += len(
-                par["values"]
-            )  # Choice parameter is one-hot encoded such that they add 1 dim for every choice
-        else:
-            dim += (
-                len(par["values"]) - 1
-            )  # Choice parameters with n_choices < 3 add n_choices - 1 dims
+    except NoOptionError:  # Look for general lb/ub array
+        _lb = config.gettensor("common", "lb")
+        _ub = config.gettensor("common", "ub")
+        bounds = torch.stack((_lb, _ub))
 
-    return dim

+    return bounds

 
 
 
[docs]def get_optimizer_options(config: Config, name: str) -> Dict[str, Any]:
diff --git a/api/_modules/aepsych/utils/index.html b/api/_modules/aepsych/utils/index.html
index 95f1d534c..403977237 100644
--- a/api/_modules/aepsych/utils/index.html
+++ b/api/_modules/aepsych/utils/index.html
@@ -25,7 +25,7 @@ 
Source code for aepsych.utils
 
 from collections.abc import Iterable
 from configparser import NoOptionError
-from typing import Any, Dict, List, Mapping, Optional, Tuple, Union
+from typing import Any, Dict, Mapping, Optional, Tuple, Union
 
 import numpy as np
 import torch
@@ -250,120 +250,35 @@ 
Source code for aepsych.utils
     return result

 
 
-def _get_ax_parameters(
-    config: Config,
-) -> Tuple[list[Dict[str, Any]], list[Dict[str, Any]], list[Dict[str, Any]]]:
-    range_parnames = config.getlist("common", "parnames", element_type=str, fallback=[])
-    lb = config.getlist("common", "lb", element_type=float, fallback=[])
-    ub = config.getlist("common", "ub", element_type=float, fallback=[])
-
-    assert (
-        len(range_parnames) == len(lb) == len(ub)
-    ), f"Length of parnames ({range_parnames}), lb ({lb}), and ub ({ub}) don't match!"
-
-    range_params = [
-        {
-            "name": parname,
-            "type": "range",
-            "value_type": config.get(str(parname), "value_type", fallback="float"),
-            "log_scale": config.getboolean(str(parname), "log_scale", fallback=False),
-            "bounds": [l, u],
-        }
-        for parname, l, u in zip(range_parnames, lb, ub)
-    ]
-
-    choice_parnames = config.getlist(
-        "common", "choice_parnames", element_type=str, fallback=[]
-    )
-    choices = [
-        config.getlist(
-            str(parname), "choices", element_type=str, fallback=["True", "False"]
+
[docs]def get_bounds(config: Config) -> torch.Tensor:
+    r"""Return the bounds for all parameters in config.
+
+    Args:
+        config (Config): The config to find the bounds from.
+
+    Returns:
+        torch.Tensor: A `[2, d]` tensor with the lower and upper bounds for each
+            parameter.
+    """
+    parnames = config.getlist("common", "parnames", element_type=str)
+
+    # Try to build a full array of bounds based on parameter-specific bounds
+    try:
+        _lower_bounds = torch.tensor(
+            [config.getfloat(par, "lower_bound") for par in parnames]
+        )
+        _upper_bounds = torch.tensor(
+            [config.getfloat(par, "upper_bound") for par in parnames]
         )
-        for parname in choice_parnames
-    ]
-    choice_params = [
-        {
-            "name": parname,
-            "type": "choice",
-            "value_type": config.get(str(parname), "value_type", fallback="str"),
-            "is_ordered": config.getboolean(str(parname), "is_ordered", fallback=False),
-            "values": choice,
-        }
-        for parname, choice in zip(choice_parnames, choices)
-    ]
-
-    fixed_parnames = config.getlist(
-        "common", "fixed_parnames", element_type=str, fallback=[]
-    )
-    values = []
-    for parname in fixed_parnames:
-        try:
-            try:
-                value: Union[float, str] = config.getfloat(str(parname), "value")
-            except ValueError:
-                value = config.get(str(parname), "value")
-
-            values.append(value)
-        except NoOptionError:
-            raise RuntimeError(f"Missing value for fixed parameter {parname}!")
-    fixed_params = [
-        {
-            "name": parname,
-            "type": "fixed",
-            "value": value,
-        }
-        for parname, value in zip(fixed_parnames, values)
-    ]
-
-    return range_params, choice_params, fixed_params
-
-
-
[docs]def get_parameters(config: Config) -> List[Dict]:
-    range_params, choice_params, fixed_params = _get_ax_parameters(config)
-    return range_params + choice_params + fixed_params

 
+        bounds = torch.stack((_lower_bounds, _upper_bounds))
 
-
[docs]def get_bounds(config: Config) -> torch.Tensor:
-    range_params, choice_params, _ = _get_ax_parameters(config)
-    # Need to sum dimensions added by both range and choice parameters
-    bounds = [parm["bounds"] for parm in range_params]
-    for par in choice_params:
-        n_vals = len(par["values"])
-        if par["is_ordered"]:
-            bounds.append(
-                [0, 1]
-            )  # Ordered choice params are encoded like continuous parameters
-        elif n_vals > 2:
-            for _ in range(n_vals):
-                bounds.append(
-                    [0, 1]
-                )  # Choice parameter is one-hot encoded such that they add 1 dim for every choice
-        else:
-            for _ in range(n_vals - 1):
-                bounds.append(
-                    [0, 1]
-                )  # Choice parameters with n_choices <= 2 add n_choices - 1 dims
-
-    return torch.tensor(bounds)

-
-
-
[docs]def get_dim(config: Config) -> int:
-    range_params, choice_params, _ = _get_ax_parameters(config)
-    # Need to sum dimensions added by both range and choice parameters
-    dim = len(range_params)  # 1 dim per range parameter
-    for par in choice_params:
-        if par["is_ordered"]:
-            dim += 1  # Ordered choice params are encoded like continuous parameters
-        elif len(par["values"]) > 2:
-            dim += len(
-                par["values"]
-            )  # Choice parameter is one-hot encoded such that they add 1 dim for every choice
-        else:
-            dim += (
-                len(par["values"]) - 1
-            )  # Choice parameters with n_choices < 3 add n_choices - 1 dims
+    except NoOptionError:  # Look for general lb/ub array
+        _lb = config.gettensor("common", "lb")
+        _ub = config.gettensor("common", "ub")
+        bounds = torch.stack((_lb, _ub))
 
-    return dim

+    return bounds

 
 
 
[docs]def get_optimizer_options(config: Config, name: str) -> Dict[str, Any]:
diff --git a/api/config.html b/api/config.html
index 0af96f7d1..428ae44dd 100644
--- a/api/config.html
+++ b/api/config.html
@@ -204,6 +204,11 @@ 
aepsych.config
+

+exception aepsych.config.ParameterConfigError[source]

+
Bases: Exception

+

 
 
 

diff --git a/api/config/index.html b/api/config/index.html
index 0af96f7d1..428ae44dd 100644
--- a/api/config/index.html
+++ b/api/config/index.html
@@ -204,6 +204,11 @@ 
aepsych.config
+

+exception aepsych.config.ParameterConfigError[source]

+
Bases: Exception

+

 
 
 

diff --git a/api/generators.html b/api/generators.html
index 9ff7b53d5..470bb48be 100644
--- a/api/generators.html
+++ b/api/generators.html
@@ -148,8 +148,7 @@ 
Submodules
    
 
  • lb (Tensor) – 
    • 
 
  • ub (Tensor) – 
    • 
-
  • points (Union[ndarray, Tensor]) – 
    • 
+
  • points (Tensor) – 
    • 
 
  • dim (Optional[int]) – 
    • 
 
  • shuffle (bool) – 
    • 
 
  • seed (Optional[int]) – 
    • 
@@ -232,29 +231,23 @@ 
    Submodules
 
    Parameters:
    
 
      
-
    • lb (Union[ndarray, Tensor]) – 
      • 
-
    • ub (Union[ndarray, Tensor]) – 
      • 
-
    • window (Union[ndarray, Tensor]) – 
      • 
-
    • points (Union[ndarray, Tensor]) – 
      • 
-
    • samples_per_point (int) – 
      • 
-
    • dim (Optional[int]) – 
      • 
-
    • shuffle (bool) – 
      • 
-
    • seed (Optional[int]) – 
      • 
+
    • points (Union[np.ndarray, torch.Tensor]) – The points that will be generated.
      • 
+
    • samples_per_point (int) – How many samples around each point to take.
      • 
+
    • dim (int, optional) – Dimensionality of the parameter space. If None, it is inferred from lb and ub.
      • 
+
    • shuffle (bool) – Whether or not to shuffle the order of the points. True by default.
      • 
+
    • seed (int, optional) – Random seed.
      • 
+
    • lb (Tensor) – 
      • 
+
    • ub (Tensor) – 
      • 
+
    • window (Union[np.ndarray, torch.Tensor]) – 
      • 
 
    
 
    
 
@@ -955,8 +948,7 @@ 
    Submodules
      
 
    • lb (Tensor) – 
      • 
 
    • ub (Tensor) – 
      • 
-
    • points (Union[ndarray, Tensor]) – 
      • 
+
    • points (Tensor) – 
      • 
 
    • dim (Optional[int]) – 
      • 
 
    • shuffle (bool) – 
      • 
 
    • seed (Optional[int]) – 
      • 
@@ -1039,29 +1031,23 @@ 
      Submodules
 
      Parameters:
      
 
        
-
      • lb (Union[ndarray, Tensor]) – 
        • 
-
      • ub (Union[ndarray, Tensor]) – 
        • 
-
      • window (Union[ndarray, Tensor]) – 
        • 
-
      • points (Union[ndarray, Tensor]) – 
        • 
-
      • samples_per_point (int) – 
        • 
-
      • dim (Optional[int]) – 
        • 
-
      • shuffle (bool) – 
        • 
-
      • seed (Optional[int]) – 
        • 
+
      • points (Union[np.ndarray, torch.Tensor]) – The points that will be generated.
        • 
+
      • samples_per_point (int) – How many samples around each point to take.
        • 
+
      • dim (int, optional) – Dimensionality of the parameter space. If None, it is inferred from lb and ub.
        • 
+
      • shuffle (bool) – Whether or not to shuffle the order of the points. True by default.
        • 
+
      • seed (int, optional) – Random seed.
        • 
+
      • lb (Tensor) – 
        • 
+
      • ub (Tensor) – 
        • 
+
      • window (Union[np.ndarray, torch.Tensor]) – 
        • 
 
      
 
      
 
diff --git a/api/generators/index.html b/api/generators/index.html
index 9ff7b53d5..470bb48be 100644
--- a/api/generators/index.html
+++ b/api/generators/index.html
@@ -148,8 +148,7 @@ 
      Submodules
@@ -1060,8 +1058,6 @@ 
      G
      
 
 
    • get_param_for() (aepsych.database.db.Database method)
 
      • 
-
    • get_parameters() (in module aepsych.utils)
-
      • 
 
    • get_raw_for() (aepsych.database.db.Database method)
 
      • 
 
    • get_replay_for() (aepsych.database.db.Database method)
@@ -1651,6 +1647,8 @@ 
      P
      
 
      • 
 
    • param_value (aepsych.database.tables.DbParamTable attribute)
 
      • 
+
    • ParameterConfigError
+
      • 
 
    • parent (aepsych.database.tables.DbConfigTable attribute)
 
        
 
      • (aepsych.database.tables.DbOutcomeTable attribute)
diff --git a/api/utils.html b/api/utils.html
index 9ed032e92..b315834d1 100644
--- a/api/utils.html
+++ b/api/utils.html
@@ -131,38 +131,22 @@ 
        aepsych.utils
-
        
-aepsych.utils.get_parameters(config)[source]
        
-
        
-
        Parameters:
        
-
        config (Config) – 
        
-
        
-
        Return type:
        
-
        List[Dict]
        
-
        
-
        
-
        
-
        
 
        
 aepsych.utils.get_bounds(config)[source]
        
-
        
+
        Return the bounds for all parameters in config.
        
+
        
 
        Parameters:
        
-
        config (Config) – 
        
+
        config (Config) – The config to find the bounds from.
        
 
        
-
        Return type:
        
-
        Tensor
        
+
        Returns:
        
+
        
+
        A [2, d] tensor with the lower and upper bounds for each
        parameter.
        
 
        
 
        
-
        
-
        
-
        
-aepsych.utils.get_dim(config)[source]
        
-
        
-
        Parameters:
        
-
        config (Config) – 
        
+
        
 
        
-
        Return type:
        
-
        int
        
+
        Return type:
        
+
        torch.Tensor
        
 
        
 
        
 
        
diff --git a/api/utils/index.html b/api/utils/index.html
index 9ed032e92..b315834d1 100644
--- a/api/utils/index.html
+++ b/api/utils/index.html
@@ -131,38 +131,22 @@ 
        aepsych.utils
-
        
-aepsych.utils.get_parameters(config)[source]
        
-
        
-
        Parameters:
        
-
        config (Config) – 
        
-
        
-
        Return type:
        
-
        List[Dict]
        
-
        
-
        
-
        
-
        
 
        
 aepsych.utils.get_bounds(config)[source]
        
-
        
+
        Return the bounds for all parameters in config.
        
+
        
 
        Parameters:
        
-
        config (Config) – 
        
+
        config (Config) – The config to find the bounds from.
        
 
        
-
        Return type:
        
-
        Tensor
        
+
        Returns:
        
+
        
+
        A [2, d] tensor with the lower and upper bounds for each
        parameter.
        
 
        
 
        
-
        
-
        
-
        
-aepsych.utils.get_dim(config)[source]
        
-
        
-
        Parameters:
        
-
        config (Config) – 
        
+
        
 
        
-
        Return type:
        
-
        int
        
+
        Return type:
        
+
        torch.Tensor
        
 
        
 
        
 
        
diff --git a/demos/ParticleEffectDemo.html b/demos/ParticleEffectDemo.html
index 0709c0c9a..65dfb3146 100644
--- a/demos/ParticleEffectDemo.html
+++ b/demos/ParticleEffectDemo.html
@@ -64,7 +64,7 @@
 
 
 
        
-
        
+
        
 
        
 
        
 
        Particle Effect Demo