refactor: rename neps.XParameter to neps.X (#149)

automl · Oct 18, 2024 · 26724bc · 26724bc
2 parents 58ee368 + efd2ec0
commit 26724bc
Show file tree

Hide file tree

Showing 54 changed files with 619 additions and 328 deletions.
diff --git a/README.md b/README.md
@@ -60,7 +60,7 @@ import logging
 
 # 1. Define a function that accepts hyperparameters and computes the validation error
 def run_pipeline(
-    hyperparameter_a: float, hyperparameter_b: int, architecture_parameter: str
+        hyperparameter_a: float, hyperparameter_b: int, architecture_parameter: str
 ) -> dict:
     # Create your model
     model = MyModel(architecture_parameter)
@@ -74,14 +74,13 @@ def run_pipeline(
 
 # 2. Define a search space of parameters; use the same parameter names as in run_pipeline
 pipeline_space = dict(
-    hyperparameter_a=neps.FloatParameter(
+    hyperparameter_a=neps.Float(
         lower=0.001, upper=0.1, log=True  # The search space is sampled in log space
     ),
-    hyperparameter_b=neps.IntegerParameter(lower=1, upper=42),
-    architecture_parameter=neps.CategoricalParameter(["option_a", "option_b"]),
+    hyperparameter_b=neps.Integer(lower=1, upper=42),
+    architecture_parameter=neps.Categorical(["option_a", "option_b"]),
 )
 
-
 # 3. Run the NePS optimization
 logging.basicConfig(level=logging.INFO)
 neps.run(

diff --git a/docs/doc_yamls/architecture_search_space.py b/docs/doc_yamls/architecture_search_space.py
@@ -86,12 +86,12 @@ def set_recursive_attribute(op_name, predecessor_values):
 
 
 pipeline_space = dict(
-    architecture=neps.ArchitectureParameter(
+    architecture=neps.Architecture(
         set_recursive_attribute=set_recursive_attribute,
         structure=structure,
         primitives=primitives,
     ),
-    optimizer=neps.CategoricalParameter(choices=["sgd", "adam"]),
-    learning_rate=neps.FloatParameter(lower=10e-7, upper=10e-3, log=True),
+    optimizer=neps.Categorical(choices=["sgd", "adam"]),
+    learning_rate=neps.Float(lower=10e-7, upper=10e-3, log=True),
 )
 
diff --git a/docs/getting_started.md b/docs/getting_started.md
@@ -35,10 +35,11 @@ In code, the usage pattern can look like this:
 import neps
 import logging
 
-def run_pipeline( # (1)!
-    hyperparameter_a: float,
-    hyperparameter_b: int,
-    architecture_parameter: str,
+
+def run_pipeline(  # (1)!
+        hyperparameter_a: float,
+        hyperparameter_b: int,
+        architecture_parameter: str,
 ) -> dict:
     # insert here your own model
     model = MyModel(architecture_parameter)
@@ -49,7 +50,7 @@ def run_pipeline( # (1)!
     )
 
     return {
-        "loss": validation_error, #! (2)
+        "loss": validation_error,  # ! (2)
         "info_dict": {
             "training_error": training_error
             # + Other metrics
@@ -58,9 +59,9 @@ def run_pipeline( # (1)!
 
 
 pipeline_space = {  # (3)!
-    "hyperparameter_b":neps.IntegerParameter(1, 42, is_fidelity=True), #! (4)
-    "hyperparameter_a":neps.FloatParameter(1e-3, 1e-1, log=True) #! (5)
-    "architecture_parameter": neps.CategoricalParameter(["option_a", "option_b", "option_c"]),
+    "hyperparameter_b": neps.Integer(1, 42, is_fidelity=True),  # ! (4)
+    "hyperparameter_a": neps.Float(1e-3, 1e-1, log=True)  # ! (5)
+    "architecture_parameter": neps.Categorical(["option_a", "option_b", "option_c"]),
 }
 
 if __name__ == "__main__":

diff --git a/docs/index.md b/docs/index.md
@@ -66,7 +66,7 @@ import logging
 
 # 1. Define a function that accepts hyperparameters and computes the validation error
 def run_pipeline(
-    hyperparameter_a: float, hyperparameter_b: int, architecture_parameter: str
+        hyperparameter_a: float, hyperparameter_b: int, architecture_parameter: str
 ) -> dict:
     # Create your model
     model = MyModel(architecture_parameter)
@@ -80,14 +80,13 @@ def run_pipeline(
 
 # 2. Define a search space of parameters; use the same parameter names as in run_pipeline
 pipeline_space = dict(
-    hyperparameter_a=neps.FloatParameter(
+    hyperparameter_a=neps.Float(
         lower=0.001, upper=0.1, log=True  # The search space is sampled in log space
     ),
-    hyperparameter_b=neps.IntegerParameter(lower=1, upper=42),
-    architecture_parameter=neps.CategoricalParameter(["option_a", "option_b"]),
+    hyperparameter_b=neps.Integer(lower=1, upper=42),
+    architecture_parameter=neps.Categorical(["option_a", "option_b"]),
 )
 
-
 # 3. Run the NePS optimization
 logging.basicConfig(level=logging.INFO)
 neps.run(

diff --git a/docs/reference/pipeline_space.md b/docs/reference/pipeline_space.md
@@ -12,10 +12,10 @@ effectively incorporate various parameter types, ensuring that NePS can utilize
 ## Parameters
 NePS currently features 4 primary hyperparameter types:
 
-* [`CategoricalParameter`][neps.search_spaces.hyperparameters.categorical.CategoricalParameter]
-* [`FloatParameter`][neps.search_spaces.hyperparameters.float.FloatParameter]
-* [`IntegerParameter`][neps.search_spaces.hyperparameters.integer.IntegerParameter]
-* [`ConstantParameter`][neps.search_spaces.hyperparameters.constant.ConstantParameter]
+* [`Categorical`][neps.search_spaces.hyperparameters.categorical.Categorical]
+* [`Float`][neps.search_spaces.hyperparameters.float.Float]
+* [`Integer`][neps.search_spaces.hyperparameters.integer.Integer]
+* [`Constant`][neps.search_spaces.hyperparameters.constant.Constant]
 
 Using these types, you can define the parameters that NePS will optimize during the search process.
 The most basic way to pass these parameters is through a Python dictionary, where each key-value
@@ -25,32 +25,32 @@ for optimizing a deep learning model:
 
 ```python
 pipeline_space = {
-    "learning_rate": neps.FloatParameter(0.00001, 0.1, log=True),
-    "num_epochs": neps.IntegerParameter(3, 30, is_fidelity=True),
-    "optimizer": neps.CategoricalParameter(["adam", "sgd", "rmsprop"]),
-    "dropout_rate": neps.ConstantParameter(0.5),
+    "learning_rate": neps.Float(0.00001, 0.1, log=True),
+    "num_epochs": neps.Integer(3, 30, is_fidelity=True),
+    "optimizer": neps.Categorical(["adam", "sgd", "rmsprop"]),
+    "dropout_rate": neps.Constant(0.5),
 }
 
-neps.run(.., pipeline_space=pipeline_space)
+neps.run(.., pipeline_space = pipeline_space)
 ```
 
 ??? example "Quick Parameter Reference"
 
-    === "`CategoricalParameter`"
+    === "`Categorical`"
 
-        ::: neps.search_spaces.hyperparameters.categorical.CategoricalParameter
+        ::: neps.search_spaces.hyperparameters.categorical.Categorical
 
-    === "`FloatParameter`"
+    === "`Float`"
 
-        ::: neps.search_spaces.hyperparameters.float.FloatParameter
+        ::: neps.search_spaces.hyperparameters.float.Float
 
-    === "`IntegerParameter`"
+    === "`Integer`"
 
-        ::: neps.search_spaces.hyperparameters.integer.IntegerParameter
+        ::: neps.search_spaces.hyperparameters.integer.Integer
 
-    === "`ConstantParameter`"
+    === "`Constant`"
 
-        ::: neps.search_spaces.hyperparameters.constant.ConstantParameter
+        ::: neps.search_spaces.hyperparameters.constant.Constant
 
 
 ## Using your knowledge, providing a Prior
@@ -70,10 +70,10 @@ import neps
 neps.run(
     ...,
     pipeline_space={
-        "learning_rate": neps.FloatParameter(1e-4, 1e-1, log=True, default=1e-2, default_confidence="medium"),
-        "num_epochs": neps.IntegerParameter(3, 30, is_fidelity=True),
-        "optimizer": neps.CategoricalParameter(["adam", "sgd", "rmsprop"], default="adam", default_confidence="low"),
-        "dropout_rate": neps.ConstantParameter(0.5),
+        "learning_rate": neps.Float(1e-4, 1e-1, log=True, default=1e-2, default_confidence="medium"),
+        "num_epochs": neps.Integer(3, 30, is_fidelity=True),
+        "optimizer": neps.Categorical(["adam", "sgd", "rmsprop"], default="adam", default_confidence="low"),
+        "dropout_rate": neps.Constant(0.5),
     }
 )
 ```

diff --git a/neps/__init__.py b/neps/__init__.py
@@ -2,28 +2,29 @@
 from neps.plot.plot import plot
 from neps.plot.tensorboard_eval import tblogger
 from neps.search_spaces import (
+    Architecture,
     ArchitectureParameter,
+    Categorical,
     CategoricalParameter,
+    Constant,
     ConstantParameter,
+    Float,
     FloatParameter,
+    Function,
     FunctionParameter,
     GraphGrammar,
+    Integer,
     IntegerParameter,
 )
 from neps.status.status import get_summary_dict, status
 
-Integer = IntegerParameter
-Float = FloatParameter
-Categorical = CategoricalParameter
-Constant = ConstantParameter
-Architecture = ArchitectureParameter
-
 __all__ = [
     "Architecture",
     "Integer",
     "Float",
     "Categorical",
     "Constant",
+    "Function",
     "ArchitectureParameter",
     "CategoricalParameter",
     "ConstantParameter",

diff --git a/neps/api.py b/neps/api.py
@@ -115,7 +115,7 @@ def run(
         >>>    validation_error = -some_parameter
         >>>    return validation_error
 
-        >>> pipeline_space = dict(some_parameter=neps.FloatParameter(lower=0, upper=1))
+        >>> pipeline_space = dict(some_parameter=neps.Float(lower=0, upper=1))
 
         >>> logging.basicConfig(level=logging.INFO)
         >>> neps.run(

diff --git a/neps/optimizers/grid_search/optimizer.py b/neps/optimizers/grid_search/optimizer.py
@@ -11,10 +11,10 @@
 from neps.optimizers.base_optimizer import BaseOptimizer, SampledConfig
 from neps.search_spaces.architecture.graph_grammar import GraphParameter
 from neps.search_spaces.domain import UNIT_FLOAT_DOMAIN
-from neps.search_spaces.hyperparameters.categorical import CategoricalParameter
-from neps.search_spaces.hyperparameters.constant import ConstantParameter
-from neps.search_spaces.hyperparameters.float import FloatParameter
-from neps.search_spaces.hyperparameters.integer import IntegerParameter
+from neps.search_spaces.hyperparameters.categorical import Categorical
+from neps.search_spaces.hyperparameters.constant import Constant
+from neps.search_spaces.hyperparameters.float import Float
+from neps.search_spaces.hyperparameters.integer import Integer
 
 if TYPE_CHECKING:
     from neps.search_spaces.search_space import SearchSpace
@@ -29,16 +29,16 @@ def _make_grid(
 ) -> list[dict[str, Any]]:
     """Get a grid of configurations from the search space.
 
-    For [`NumericalParameter`][neps.search_spaces.NumericalParameter] hyperparameters,
+    For [`Numerical`][neps.search_spaces.Numerical] hyperparameters,
     the parameter `size_per_numerical_hp=` is used to determine a grid. If there are
     any duplicates, e.g. for an
-    [`IntegerParameter`][neps.search_spaces.IntegerParameter], then we will
+    [`Integer`][neps.search_spaces.Integer], then we will
     remove duplicates.
 
-    For [`CategoricalParameter`][neps.search_spaces.CategoricalParameter]
+    For [`Categorical`][neps.search_spaces.Categorical]
     hyperparameters, we include all the choices in the grid.
 
-    For [`ConstantParameter`][neps.search_spaces.ConstantParameter] hyperparameters,
+    For [`Constant`][neps.search_spaces.Constant] hyperparameters,
     we include the constant value in the grid.
 
     !!! note "TODO"
@@ -65,11 +65,11 @@ def _make_grid(
             # If this is resolved, please update the docstring!
             case GraphParameter():
                 raise ValueError("Trying to create a grid for graphs!")
-            case CategoricalParameter():
+            case Categorical():
                 param_ranges[name] = list(hp.choices)
-            case ConstantParameter():
+            case Constant():
                 param_ranges[name] = [hp.value]
-            case IntegerParameter() | FloatParameter():
+            case Integer() | Float():
                 if hp.is_fidelity:
                     param_ranges[name] = [hp.upper]
                     continue

diff --git a/neps/optimizers/multi_fidelity/ifbo.py b/neps/optimizers/multi_fidelity/ifbo.py
@@ -19,7 +19,7 @@
 from neps.sampling.samplers import Sampler
 from neps.search_spaces.domain import Domain
 from neps.search_spaces.encoding import CategoricalToUnitNorm, ConfigEncoder
-from neps.search_spaces.search_space import FloatParameter, IntegerParameter, SearchSpace
+from neps.search_spaces.search_space import Float, Integer, SearchSpace
 
 if TYPE_CHECKING:
     from neps.state.optimizer import BudgetInfo
@@ -48,7 +48,7 @@ def _adjust_pipeline_space_to_match_stepsize(
     fidelity = pipeline_space.fidelity
     fidelity_name = pipeline_space.fidelity_name
     assert fidelity_name is not None
-    assert isinstance(fidelity, FloatParameter | IntegerParameter)
+    assert isinstance(fidelity, Float | Integer)
     if fidelity.log:
         raise NotImplementedError("Log fidelity not yet supported")
 

diff --git a/neps/optimizers/multi_fidelity/successive_halving.py b/neps/optimizers/multi_fidelity/successive_halving.py
@@ -20,10 +20,10 @@
     RandomUniformPolicy,
 )
 from neps.search_spaces import (
-    CategoricalParameter,
-    ConstantParameter,
-    FloatParameter,
-    IntegerParameter,
+    Categorical,
+    Constant,
+    Float,
+    Integer,
     SearchSpace,
 )
 
@@ -34,12 +34,10 @@
 
 logger = logging.getLogger(__name__)
 
-CUSTOM_FLOAT_CONFIDENCE_SCORES = dict(FloatParameter.DEFAULT_CONFIDENCE_SCORES)
+CUSTOM_FLOAT_CONFIDENCE_SCORES = dict(Float.DEFAULT_CONFIDENCE_SCORES)
 CUSTOM_FLOAT_CONFIDENCE_SCORES.update({"ultra": 0.05})
 
-CUSTOM_CATEGORICAL_CONFIDENCE_SCORES = dict(
-    CategoricalParameter.DEFAULT_CONFIDENCE_SCORES
-)
+CUSTOM_CATEGORICAL_CONFIDENCE_SCORES = dict(Categorical.DEFAULT_CONFIDENCE_SCORES)
 CUSTOM_CATEGORICAL_CONFIDENCE_SCORES.update({"ultra": 8})
 
 
@@ -187,7 +185,7 @@ def _get_rung_map(self, s: int = 0) -> dict:
         for i in reversed(range(nrungs)):
             rung_map[i + s] = (
                 int(_max_budget)
-                if isinstance(self.pipeline_space.fidelity, IntegerParameter)
+                if isinstance(self.pipeline_space.fidelity, Integer)
                 else _max_budget
             )
             _max_budget /= self.eta
@@ -465,15 +463,15 @@ def _enhance_priors(self, confidence_score: dict[str, float] | None = None) -> N
             return
 
         for k, v in self.pipeline_space.items():
-            if v.is_fidelity or isinstance(v, ConstantParameter):
+            if v.is_fidelity or isinstance(v, Constant):
                 continue
-            if isinstance(v, FloatParameter | IntegerParameter):
+            if isinstance(v, Float | Integer):
                 if confidence_score is None:
                     confidence = CUSTOM_FLOAT_CONFIDENCE_SCORES[self.prior_confidence]
                 else:
                     confidence = confidence_score["numeric"]
                 self.pipeline_space[k].default_confidence_score = confidence
-            elif isinstance(v, CategoricalParameter):
+            elif isinstance(v, Categorical):
                 if confidence_score is None:
                     confidence = CUSTOM_CATEGORICAL_CONFIDENCE_SCORES[
                         self.prior_confidence