Remove nptyping reliance, Github workflows, updated CirleCI runs

.circleci/config.yml

@@ -3,6 +3,7 @@ version: 2
 shared: &shared
   working_directory: ~/finsim
 
+  resource_class: medium+
   steps:
     - checkout
 

.github/workflow/tests.yaml → .github/workflows/tests.yaml

@@ -2,9 +2,10 @@ name: Python package CI
 
 on:
   push:
-    branches: [ master ]
+    branches:
+      - '**'
   pull_request:
-    branches: [ master ]
+    branches: [ master, develop ]
 
 jobs:
   build_and_test:
@@ -23,9 +24,9 @@ jobs:
       - name: Install dependencies
         run: |
           pip install --user --upgrade pip
-          pip install --user --upgrade setuptools wheel Cython
+          pip install --user --upgrade setuptools wheel
           pip cache purge
           pip install --user .
           pip install --user .[test]
       - name: Run tests
-        run: pytest --doctest-cython
+        run: pytest

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "finsim"
-version = "1.2.0"
+version = "1.2.1"
 authors = [
     {name = "Kwan Yuet Stephen Ho", email = "stephenhky@yahoo.com.hk"}
 ]
@@ -37,7 +37,6 @@ dependencies = [
     "openpyxl>=3.1.0",
     "numba",
     "typing-extensions",
-    "nptyping",
     "pydantic"
 ]
 

src/finsim/estimate/fit.py

@@ -1,9 +1,9 @@
 
-from typing import Literal
+from typing import Literal, Annotated
 from itertools import product
 
 import numpy as np
-from nptyping import NDArray, Shape, Float, Datetime64
+from numpy.typing import NDArray
 
 from .constants import dividing_factors_dict
 from .native.pyfit import python_fit_BlackScholesMerton_model, python_fit_multivariate_BlackScholesMerton_model
@@ -12,8 +12,8 @@
 # Note: always round-off to seconds first, but flexible about the unit to be used.
 
 def fit_BlackScholesMerton_model(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        prices: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year'
 ) -> tuple[float, float]:
     """Fit a Black-Scholes-Merton model to price data to estimate rate of return and volatility.
@@ -46,10 +46,10 @@ def fit_BlackScholesMerton_model(
 
 
 def fit_multivariate_BlackScholesMerton_model(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        multiprices: NDArray[Shape["*, *"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        multiprices: Annotated[NDArray[np.float64], Literal["1D array"]],
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year',
-) -> tuple[NDArray[Shape["*"], Float], NDArray[Shape["*, *"], Float]]:
+) -> tuple[Annotated[NDArray[np.float64], Literal["1D array"]], Annotated[NDArray[np.float64], Literal["2D array"]]]:
     """Fit a multivariate Black-Scholes-Merton model to price data for multiple assets.
 
     This function estimates the parameters of the multivariate Black-Scholes-Merton model,
@@ -82,9 +82,9 @@ def fit_multivariate_BlackScholesMerton_model(
 ######## routines below are for time-weighted portfolio building
 
 def fit_timeweighted_BlackScholesMerton_model(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        prices: NDArray[Shape["*"], Float],
-        weights: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year'
 ) -> tuple[float, float]:
     """Fit a time-weighted Black-Scholes-Merton model to price data.
@@ -120,11 +120,11 @@ def fit_timeweighted_BlackScholesMerton_model(
 
 
 def fit_timeweighted_multivariate_BlackScholesMerton_model(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        multiprices: NDArray[Shape["*, *"], Float],
-        weights: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        multiprices: Annotated[NDArray[np.float64], Literal["2D array"]],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year'
-) -> tuple[NDArray[Shape["*"], Float], NDArray[Shape["*, *"], Float]]:
+) -> tuple[Annotated[NDArray[np.float64], Literal["1D array"]], Annotated[NDArray[np.float64], Literal["2D array"]]]:
     """Fit a time-weighted multivariate Black-Scholes-Merton model to price data for multiple assets.
 
     This function estimates the parameters of the multivariate Black-Scholes-Merton model

src/finsim/estimate/native/pyfit.py

@@ -1,13 +1,15 @@
 
+from typing import Literal, Annotated
+
 import numpy as np
+from numpy.typing import NDArray
 import numba as nb
-from nptyping import NDArray, Shape, Float
 
 
 @nb.njit
 def python_fit_BlackScholesMerton_model(
-        ts: NDArray[Shape["*"], Float],
-        prices: NDArray[Shape["*"], Float]
+        ts: Annotated[NDArray[np.float64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]]
 ) -> tuple[float, float]:
     """Fit a Black-Scholes-Merton model to price data using Python implementation.
 
@@ -35,9 +37,9 @@ def python_fit_BlackScholesMerton_model(
 
 @nb.njit
 def python_fit_multivariate_BlackScholesMerton_model(
-        ts: NDArray[Shape["*"], Float],
-        multiprices: NDArray[Shape["*, *"], Float]
-) -> tuple[NDArray[Shape["*"], Float], NDArray[Shape["*, *"], Float]]:
+        ts: Annotated[NDArray[np.float64], Literal["1D array"]],
+        multiprices: Annotated[NDArray[np.float64], Literal["2D array"]]
+) -> tuple[Annotated[NDArray[np.float64], Literal["1D array"]], Annotated[NDArray[np.float64], Literal["2D array"]]]:
     """Fit a multivariate Black-Scholes-Merton model to price data for multiple assets.
 
     This function estimates the parameters of the multivariate Black-Scholes-Merton model,

src/finsim/estimate/native/pyrisk.py

@@ -1,13 +1,15 @@
 
+from typing import Literal, Annotated
+
 import numpy as np
+from numpy.typing import NDArray
 import numba as nb
-from nptyping import NDArray, Shape, Float
 
 
 @nb.njit
 def python_estimate_downside_risk(
-        ts: NDArray[Shape["*"], Float],
-        prices: NDArray[Shape["*"], Float],
+        ts: Annotated[NDArray[np.float64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         target_return: float
 ) -> float:
     """Estimate the downside risk of an asset based on historical price data.
@@ -35,8 +37,8 @@ def python_estimate_downside_risk(
 
 @nb.njit
 def python_estimate_upside_risk(
-        ts: NDArray[Shape["*"], Float],
-        prices: NDArray[Shape["*"], Float],
+        ts: Annotated[NDArray[np.float64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         target_return: float
 ) -> float:
     """Estimate the upside risk of an asset based on historical price data.

src/finsim/estimate/risk.py

@@ -1,17 +1,17 @@
 
-from typing import Literal
+from typing import Literal, Annotated
 
 import numpy as np
+from numpy.typing import NDArray
 from scipy import stats
-from nptyping import NDArray, Shape, Float, Datetime64
 
 from .native.pyrisk import python_estimate_downside_risk, python_estimate_upside_risk
 from .constants import dividing_factors_dict
 
 
 def estimate_downside_risk(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        prices: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         target_return: float,
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year'
 ) -> float:
@@ -43,8 +43,8 @@ def estimate_downside_risk(
 
 
 def estimate_upside_risk(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        prices: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         target_return: float,
         unit: Literal['second', 'minute', 'hour', 'day', 'year'] = 'year'
 ) -> float:
@@ -76,9 +76,9 @@ def estimate_upside_risk(
 
 
 def estimate_beta(
-        timestamps: NDArray[Shape["*"], Datetime64],
-        prices: NDArray[Shape["*"], Float],
-        market_prices: NDArray[Shape["*"], Float],
+        timestamps: Annotated[NDArray[np.datetime64], Literal["1D array"]],
+        prices: Annotated[NDArray[np.float64], Literal["1D array"]],
+        market_prices: Annotated[NDArray[np.float64], Literal["1D array"]],
         unit: Literal['second', 'minute', 'hour', 'day', 'year']='year'
 ) -> float:
     """Estimate the beta coefficient of an asset relative to market performance.

src/finsim/portfolio/optimize/metrics.py

@@ -1,15 +1,16 @@
 
-from typing import Literal
+from typing import Literal, Annotated
 
-from nptyping import NDArray, Shape, Float
+import numpy as np
+from numpy.typing import NDArray
 
 from .native.pymetrics import python_sharpe_ratio, python_mpt_costfunction, python_mpt_entropy_costfunction
 
 
 def sharpe_ratio(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float
 ) -> float:
     """Calculate the Sharpe ratio for a portfolio.
@@ -27,9 +28,9 @@ def sharpe_ratio(
 
 
 def mpt_costfunction(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float,
         lamb: float,
         V0: float=10.
@@ -52,9 +53,9 @@ def mpt_costfunction(
 
 
 def mpt_entropy_costfunction(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float,
         lamb0: float,
         lamb1: float,

src/finsim/portfolio/optimize/native/pymetrics.py

@@ -1,14 +1,16 @@
 
+from typing import Literal, Annotated
+
 import numpy as np
+from numpy.typing import NDArray
 import numba as nb
-from nptyping import NDArray, Shape, Float
 
 
 @nb.njit
 def python_sharpe_ratio(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float
 ) -> float:
     """Calculate the Sharpe ratio for a portfolio using Python/Numba.
@@ -30,9 +32,9 @@ def python_sharpe_ratio(
 
 @nb.njit
 def python_mpt_costfunction(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float,
         lamb: float,
         V0: float=10.
@@ -57,9 +59,9 @@ def python_mpt_costfunction(
 
 @nb.njit
 def python_mpt_entropy_costfunction(
-        weights: NDArray[Shape["*"], Float],
-        r: NDArray[Shape["*"], Float],
-        cov: NDArray[Shape["*, *"], Float],
+        weights: Annotated[NDArray[np.float64], Literal["1D array"]],
+        r: Annotated[NDArray[np.float64], Literal["1D array"]],
+        cov: Annotated[NDArray[np.float64], Literal["2D array"]],
         rf: float,
         lamb0: float,
         lamb1: float,