fix: drop pysen

.flexci/config.pbtxt

@@ -4,7 +4,7 @@ configs {
     requirement {
       cpu: 2
       gpu: 1
-      memory: 20
+      memory: 30
       disk: 10
     }
     time_limit: {
@@ -20,7 +20,7 @@ configs {
     requirement {
       cpu: 2
       gpu: 1
-      memory: 20
+      memory: 30
       disk: 10
     }
     time_limit: {
@@ -36,7 +36,7 @@ configs {
     requirement {
       cpu: 2
       gpu: 1
-      memory: 20
+      memory: 30
       disk: 10
     }
     time_limit: {

Makefile

@@ -24,7 +24,7 @@ test-cov:
 	$(RUN) pytest --cov=$(PROJECT_NAME) --cov-report=xml -m "not gpu"
 
 .PHONY: lint
-lint: lint-pysen
+lint: lint-black lint-isort flake8 mypy
 
 .PHONY: lint-black
 lint-black:
@@ -40,27 +40,19 @@ mypy:
 
 .PHONY: flake8
 flake8:
-	$(RUN) flake8 $(PROJECT_NAME)
+	$(RUN) pflake8 $(PROJECT_NAME)
 
 .PHONY: format
-format: format-pysen
+format: format-black format-isort
 
 .PHONY: format-black
 format-black:
-	$(RUN) black --quiet --skip-magic-trailing-comma .
+	$(RUN) black --quiet .
 
 .PHONY: format-isort
 format-isort:
 	$(RUN) isort --force-single-line-imports --quiet .
 
-.PHONY: format-pysen
-format-pysen:
-	$(RUN) pysen run format
-
-.PHONY: lint-pysen
-lint-pysen:
-	$(RUN) pysen run lint
-
 .PHONY: doc
 doc:
 	@cd docs && make html

examples/example_hedging_variance_swap.py

@@ -26,7 +26,7 @@
     total_vega = torch.zeros_like(spot)
     for option, strike in zip(options_list, strikes_list):
         lm = (spot / strike).log()
-        vega = BlackScholes(option).vega(lm, t, v) / (strike**2)
+        vega = BlackScholes(option).vega(lm, t, v) / (strike ** 2)
         total_vega += vega
         if option.call:
             # 2 is for call and put

pfhedge/features/_base.py

@@ -74,7 +74,7 @@ def is_state_dependent(self) -> bool:
         # If a feature uses the state of a hedger, it is state dependent.
         return getattr(self, "hedger") is not None
 
-    def __str__(self):
+    def __str__(self) -> str:
         return self.name
 
     # TODO(simaki) Remove later

pfhedge/features/_getter.py

@@ -14,7 +14,7 @@ class FeatureFactory:
     _features: Dict[str, Type[Feature]]
 
     # singleton
-    def __new__(cls, *args, **kwargs):
+    def __new__(cls, *args: Any, **kwargs: Any) -> "FeatureFactory":
         if not hasattr(cls, "_instance"):
             cls._instance = super().__new__(cls)
             cls._instance._features = OrderedDict()

pfhedge/features/features.py

@@ -134,7 +134,7 @@ def __str__(self) -> str:
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if isinstance(time_step, int) else ...
-        output = self.derivative.ul().spot[:, index].unsqueeze(-1)
+        output = self.derivative.ul().spot[:, index].unsqueeze(-1)  # type: ignore
         if self.log:
             output.log_()
         return output
@@ -147,7 +147,7 @@ class UnderlierLogSpot(UnderlierSpot):
         ``'underlier_log_spot'``
     """
 
-    def __init__(self):
+    def __init__(self) -> None:
         super().__init__(log=True)
 
 
@@ -167,7 +167,7 @@ def __str__(self) -> str:
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if isinstance(time_step, int) else ...
-        output = self.derivative.spot[:, index].unsqueeze(-1)
+        output = self.derivative.spot[:, index].unsqueeze(-1)  # type: ignore
         if self.log:
             output.log_()
         return output
@@ -186,7 +186,7 @@ class Volatility(StateIndependentFeature):
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if isinstance(time_step, int) else ...
-        return self.derivative.ul().volatility[:, index].unsqueeze(-1)
+        return self.derivative.ul().volatility[:, index].unsqueeze(-1)  # type: ignore
 
 
 class Variance(StateIndependentFeature):
@@ -200,7 +200,7 @@ class Variance(StateIndependentFeature):
 
     def get(self, time_step: Optional[int]) -> Tensor:
         index = [time_step] if isinstance(time_step, int) else ...
-        return self.derivative.ul().variance[:, index].unsqueeze(-1)
+        return self.derivative.ul().variance[:, index].unsqueeze(-1)  # type: ignore
 
 
 class PrevHedge(Feature):
@@ -305,7 +305,7 @@ class Zeros(StateIndependentFeature):
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if time_step is not None else ...
-        return torch.zeros_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)
+        return torch.zeros_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)  # type: ignore
 
 
 class Ones(StateIndependentFeature):
@@ -335,7 +335,7 @@ class Ones(StateIndependentFeature):
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if time_step is not None else ...
-        return torch.ones_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)
+        return torch.ones_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)  # type: ignore
 
 
 class Empty(StateIndependentFeature):
@@ -365,7 +365,7 @@ class Empty(StateIndependentFeature):
 
     def get(self, time_step: Optional[int] = None) -> Tensor:
         index = [time_step] if time_step is not None else ...
-        return torch.empty_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)
+        return torch.empty_like(self.derivative.ul().spot[..., index]).unsqueeze(-1)  # type: ignore
 
 
 class MaxMoneyness(StateIndependentFeature):

pfhedge/instruments/derivative/base.py

@@ -305,7 +305,7 @@ def moneyness(self, time_step: Optional[int] = None, log: bool = False) -> Tenso
             torch.Tensor
         """
         index = ... if time_step is None else [time_step]
-        output = self.underlier.spot[..., index] / self.strike
+        output = self.underlier.spot[..., index] / self.strike  # type: ignore
         if log:
             output = output.log()
         return output
@@ -389,7 +389,7 @@ def max_log_moneyness(self, time_step: Optional[int] = None) -> Tensor:
 class BaseOption(BaseDerivative, OptionMixin):
     """(deprecated) Base class for options."""
 
-    def __init__(self):
+    def __init__(self) -> None:
         super().__init__()
         raise DeprecationWarning(
             "BaseOption is deprecated. Inherit `BaseDerivative` and `OptionMixin` instead."

pfhedge/nn/modules/bs/american_binary.py

@@ -82,7 +82,9 @@ def __init__(
         self.derivative = derivative
 
     @classmethod
-    def from_derivative(cls, derivative):
+    def from_derivative(
+        cls, derivative: AmericanBinaryOption
+    ) -> "BSAmericanBinaryOption":
         """Initialize a module from a derivative.
 
         Args:

pfhedge/nn/modules/bs/black_scholes.py

@@ -1,4 +1,5 @@
 from collections import OrderedDict
+from typing import Any
 from typing import Callable
 from typing import Dict
 from typing import Iterator
@@ -17,7 +18,7 @@ class BlackScholesModuleFactory:
     _modules: Dict[str, Type[Module]]
 
     # singleton
-    def __new__(cls, *args, **kwargs):
+    def __new__(cls, *args: Any, **kwargs: Any) -> "BlackScholesModuleFactory":
         if not hasattr(cls, "_instance"):
             cls._instance = super().__new__(cls)
             cls._instance._modules = OrderedDict()
@@ -97,5 +98,5 @@ class BlackScholes(Module):
     vega: Callable[..., Tensor]  # vega(self, ...) -> Tensor
     theta: Callable[..., Tensor]  # theta(self, ...) -> Tensor
 
-    def __new__(cls, derivative):
-        return BlackScholesModuleFactory().get_class_from_derivative(derivative)
+    def __new__(cls, derivative: "Derivative") -> "BlackScholes":
+        return BlackScholesModuleFactory().get_class_from_derivative(derivative)  # type: ignore

pfhedge/nn/modules/bs/european.py

@@ -73,7 +73,7 @@ def __init__(
         self.derivative = derivative
 
     @classmethod
-    def from_derivative(cls, derivative):
+    def from_derivative(cls, derivative: EuropeanOption) -> "BSEuropeanOption":
         """Initialize a module from a derivative.
 
         Args:

pfhedge/nn/modules/bs/european_binary.py

@@ -75,7 +75,9 @@ def __init__(
         self.derivative = derivative
 
     @classmethod
-    def from_derivative(cls, derivative):
+    def from_derivative(
+        cls, derivative: EuropeanBinaryOption
+    ) -> "BSEuropeanBinaryOption":
         """Initialize a module from a derivative.
 
         Args:

pfhedge/nn/modules/bs/lookback.py

@@ -83,7 +83,7 @@ def __init__(
         self.derivative = derivative
 
     @classmethod
-    def from_derivative(cls, derivative):
+    def from_derivative(cls, derivative: LookbackOption) -> "BSLookbackOption":
         """Initialize a module from a derivative.
 
         Args:

pfhedge/nn/modules/hedger.py

@@ -475,7 +475,7 @@ def compute_loss(
         """
         with torch.set_grad_enabled(enable_grad):
 
-            def _get_loss():
+            def _get_loss() -> Tensor:
                 derivative.simulate(n_paths=n_paths, init_state=init_state)
                 portfolio = self.compute_portfolio(derivative, hedge=hedge)
                 return self.criterion(portfolio, derivative.payoff())
@@ -666,7 +666,7 @@ def price(
         """
         with torch.set_grad_enabled(enable_grad):
 
-            def _get_price():
+            def _get_price() -> Tensor:
                 derivative.simulate(n_paths=n_paths, init_state=init_state)
                 portfolio = self.compute_portfolio(derivative, hedge)
                 # Negative because selling

pfhedge/nn/modules/loss.py

@@ -1,4 +1,5 @@
 from abc import ABC
+from abc import abstractmethod
 from typing import Callable
 
 import torch
@@ -20,6 +21,7 @@
 class HedgeLoss(Module, ABC):
     """Base class for hedging criteria."""
 
+    @abstractmethod
     def forward(self, input: Tensor, target: TensorOrScalar = 0.0) -> Tensor:
         """Returns the loss of the profit-loss distribution.
 
@@ -39,6 +41,7 @@ def forward(self, input: Tensor, target: TensorOrScalar = 0.0) -> Tensor:
         Returns:
             torch.Tensor
         """
+        pass
 
     def cash(self, input: Tensor, target: TensorOrScalar = 0.0) -> Tensor:
         """Returns the cash amount which is as preferable as

pfhedge/stochastic/rough_bergomi.py

@@ -122,9 +122,7 @@ def discrete_TBSS_fn(k: torch.Tensor, a: TensorOrScalar) -> torch.Tensor:
     _gamma = torch.cat([torch.zeros(2, dtype=dtype, device=device), _gamma], dim=0)
     _Xi = dW1[:, :, 0]
     _GXi_convolve = torch.nn.functional.conv1d(
-        _gamma.flip(0)[None, None, :],
-        _Xi[:, None, :],
-        padding=_Xi.size(1) - 1,
+        _gamma.flip(0)[None, None, :], _Xi[:, None, :], padding=_Xi.size(1) - 1
     )[0, :, :]
     _Y2 = _GXi_convolve[:, -n_steps:].flip(1)
     Y = torch.sqrt(2 * alpha_tensor + 1) * (_Y1 + _Y2)