sovai 0.2.60__tar.gz → 0.2.62__tar.gz

{sovai-0.2.60 → sovai-0.2.62}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: sovai
-Version: 0.2.60
+Version: 0.2.62
 Summary: python for asset management
 License: Proprietary License
 Author: Derek Snow

{sovai-0.2.60 → sovai-0.2.62}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [project]
 name = "sovai"
-version = "0.2.60"
+version = "0.2.62"
 description = "python for asset management"
 authors = [
     {name = "Derek Snow", email = "d.snow@outlook.com"}

{sovai-0.2.60 → sovai-0.2.62}/sovai/extensions/pandas_extensions.py

@@ -1,7 +1,6 @@
 import pandas as pd
 import numpy as np
-import plotly.express as px
-from numpy.linalg import norm
+
 
 # from extensions.pfa_feature_selector import PFA ## # ## ### 
 
@@ -11,72 +10,6 @@ pd.options.display.float_format = "{:.3f}".format
 
 # from sovai.extensions.shapley_importance import run_simulations_frame
 
-from sovai.extensions.feature_neutralizer import (
-    neutralize_features_function,
-    orthogonalize_features_function,
-)
-
-from sovai.extensions.fractional_differencing import fractional_diff
-
-from sovai.extensions.anomalies import (
-    anomaly_scores,
-    anomaly_global,
-    anomaly_local,
-    anomaly_cluster,
-    anomaly_reconstruction,
-)
-
-from sovai.extensions.pairwise import distance_calc, relative_distance_calc
-
-from sovai.extensions.clustering import (
-    cluster,
-    cluster_summary,
-    feature_cent,
-    vizualisation_cluster,
-    vizualisation_scatter,
-    vizualisation_animation,
-)
-
-from sovai.extensions.nowcasting import nowcast_data_source, nowcast_plot_source
-
-from sovai.extensions.change_point_generator import (
-    perform_cusum_analysis,
-    run_cusum_dashboard,
-)
-
-from sovai.extensions.regime_change import (
-    perform_regime_change_analysis,
-    run_regime_change_dashboard,
-)
-
-from sovai.extensions.regime_change_pca import (
-    perform_pca_regime_change_analysis,
-    run_pca_regime_change_dashboard,
-)
-
-from sovai.extensions.time_decomposition import (
-    perform_comprehensive_analysis,
-    run_comprehensive_analysis_dashboard,
-)
-
-from sovai.extensions.feature_extraction import feature_extractor
-
-from sovai.extensions.dimensionality_reduction import dimensionality_reduction
-
-from sovai.extensions.feature_importance import (
-    random_projection_importance,
-    fast_nonlinear_diverse_selector,
-    fast_ica_selector,
-    truncated_svd_selector,
-    sparse_random_projection_selector,
-)
-
-from sovai.extensions.weight_optimization import WeightOptimization
-
-from sovai.extensions.signal_evaluation import SignalEvaluator
-
-from sovai.extensions.technical_indicators import techn_indicators
-
 
 import warnings
 
@@ -98,8 +31,8 @@ import pandas as pd
 
 import pandas as pd
 from typing import Union, List
-from sovai.extensions.ask_df_llm import Ask
-from sovai.extensions.filter_df import Filter
+# from sovai.extensions.ask_df_llm import Ask
+
 import pandas as pd
 from functools import lru_cache
 import re
@@ -124,6 +57,9 @@ class CustomDataFrame(pd.DataFrame):
         :param verbose: If True, print detailed information about the filtering process.
         :return: A filtered CustomDataFrame.
         """
+
+        from sovai.extensions.filter_df import Filter
+
         filtered_df = Filter.filter(conditions, df_multi=self, verbose=verbose)
         return CustomDataFrame(filtered_df)
     
@@ -230,6 +166,7 @@ class CustomDataFrame(pd.DataFrame):
             return CustomDataFrame(latest_data)
 
     def plot_line(self, column=None, tickers=None, n=50):
+        import plotly.express as px
         # Your plot_line implementation
         if isinstance(self.index, pd.MultiIndex):
             df = self.reset_index()
@@ -308,6 +245,9 @@ class CustomDataFrame(pd.DataFrame):
         metric="pearson",
         calculations=["mean"],
     ):
+        
+        from sovai.extensions.pairwise import distance_calc
+
         result = distance_calc(
             self,
             orient=orient,
@@ -326,6 +266,9 @@ class CustomDataFrame(pd.DataFrame):
         metric="pearson",
         calculations=["mean"],
     ):
+        
+        from sovai.extensions.pairwise import distance_calc, relative_distance_calc
+
         result = relative_distance_calc(
             self,
             orient=orient,
@@ -337,6 +280,15 @@ class CustomDataFrame(pd.DataFrame):
         return CustomDataFrame(result)
 
     def cluster(self, orient="cluster", features=None):
+
+        from sovai.extensions.clustering import (
+            cluster,
+            cluster_summary,
+            feature_cent,
+            vizualisation_cluster,
+            vizualisation_scatter,
+            vizualisation_animation,
+        )
         if orient == "cluster":
             result = cluster(self, features)
         elif orient == "summary":
@@ -358,10 +310,10 @@ class CustomDataFrame(pd.DataFrame):
 
     import numpy as np
     import pandas as pd
-    from numpy.linalg import norm
 
 
     def cointegration(self, on="ticker", shift=12):
+
         """
         Calculate an approximate cointegration proxy using shifted cosine similarity.
 
@@ -370,6 +322,9 @@ class CustomDataFrame(pd.DataFrame):
         :param shift: The number of periods to shift for lagged comparison (default 1 month).
         :return: DataFrame of shifted cosine similarities.
         """
+
+        from numpy.linalg import norm
+
         # Shift the DataFrame by the specified lag
         shifted_df = self.groupby(level=on).shift(-shift)
 
@@ -462,6 +417,8 @@ class CustomDataFrame(pd.DataFrame):
     #     return CustomDataFrame(avg_shap_values)
 
     def orthogonalize_features(self, method="gram_schmidt"):
+        from sovai.extensions.feature_neutralizer import orthogonalize_features_function
+        
         return CustomDataFrame(orthogonalize_features_function(self, method))
 
     def importance_reorder(self):
@@ -469,6 +426,7 @@ class CustomDataFrame(pd.DataFrame):
 
 
     def neutralize_features(self, method="pca"):
+        from sovai.extensions.feature_neutralizer import neutralize_features_function
         return CustomDataFrame(neutralize_features_function(self, method))
 
     def ticker(self, ticker="AAPL"):
@@ -710,6 +668,8 @@ class CustomDataFrame(pd.DataFrame):
         return CustomDataFrame(df_merge)
 
     def fractional_difference(self, d=0.845, m=100, groupby_column="ticker"):
+        from sovai.extensions.fractional_differencing import fractional_diff
+
         """
         Apply fractional differencing to each group for all columns, or to the entire DataFrame if groupby_column is not present.
         :param d: Differencing parameter.
@@ -740,6 +700,14 @@ class CustomDataFrame(pd.DataFrame):
 
     def anomalies(self, method="scores", ticker=None):
 
+        from sovai.extensions.anomalies import (
+            anomaly_scores,
+            anomaly_global,
+            anomaly_local,
+            anomaly_cluster,
+            anomaly_reconstruction,
+        )
+
         if method == "scores":
             return CustomDataFrame(anomaly_scores(self, ticker=ticker))
         elif method == "global":
@@ -759,6 +727,15 @@ class CustomDataFrame(pd.DataFrame):
 
         from sovai.extensions.shapley_global_importance import run_simulations_global_importance
 
+        from sovai.extensions.feature_importance import (
+            random_projection_importance,
+            fast_nonlinear_diverse_selector,
+            fast_ica_selector,
+            truncated_svd_selector,
+            sparse_random_projection_selector,
+        )
+
+
         if method == "random_projection":
             return random_projection_importance(self)
         elif method == "fourier":
@@ -781,15 +758,26 @@ class CustomDataFrame(pd.DataFrame):
 
     def nowcast_data(self, feature=None):
 
+        from sovai.extensions.nowcasting import nowcast_data_source
+
+
         return CustomDataFrame(nowcast_data_source(self, selected_feature=feature))
 
 
     def nowcast_plot(self, feature=None):
 
+        from sovai.extensions.nowcasting import nowcast_plot_source
+
         return nowcast_plot_source(self, feature=feature)
 
 
     def change_point(self, method="data", ticker=None, feature=None):
+
+        from sovai.extensions.change_point_generator import (
+        perform_cusum_analysis,
+        run_cusum_dashboard,
+            )
+
         if ticker is None or feature is None:
             print("Ticker and/or feature not provided. We will choose on your behalf.")
             if ticker is None:
@@ -819,15 +807,26 @@ class CustomDataFrame(pd.DataFrame):
                 print(f"Selected feature: {feature}")
 
         if method == "data":
+        
+            from sovai.extensions.regime_change import perform_regime_change_analysis
+
             analysis_df = perform_regime_change_analysis(self, ticker, feature)
             result = CustomDataFrame(analysis_df)
             return result
         elif method == "plot":
+
+            from sovai.extensions.regime_change import run_regime_change_dashboard
             return run_regime_change_dashboard(self, ticker=ticker, feature=feature)
         else:
             raise ValueError("Invalid method. Choose from 'data', 'plot'")
 
     def pca_regime_change(self, method="data", ticker=None):
+
+
+        from sovai.extensions.regime_change_pca import (
+            perform_pca_regime_change_analysis,
+            run_pca_regime_change_dashboard,
+        )
         if ticker is None:
             print("Ticker not provided. We will choose on your behalf.")
             ticker = self.index.get_level_values("ticker").unique()[0]
@@ -841,6 +840,13 @@ class CustomDataFrame(pd.DataFrame):
             raise ValueError("Invalid method. Choose from 'data', 'plot'")
 
     def time_decomposition(self, method="data", ticker=None, feature=None):
+
+
+        from sovai.extensions.time_decomposition import (
+            perform_comprehensive_analysis,
+            run_comprehensive_analysis_dashboard,
+        )
+
         if ticker is None or feature is None:
             print("Ticker and/or feature not provided. We will choose on your behalf.")
             if ticker is None:
@@ -873,6 +879,8 @@ class CustomDataFrame(pd.DataFrame):
         """
         Extracts features from the CustomDataFrame and returns a new CustomDataFrame with the extracted features.
         """
+        from sovai.extensions.feature_extraction import feature_extractor
+
         result_pd = feature_extractor(
             self, entity_col, date_col, lookback, features, every, verbose
         )
@@ -894,6 +902,12 @@ class CustomDataFrame(pd.DataFrame):
         CustomDataFrame: Reduced data in panel format
         """
 
+
+
+        from sovai.extensions.dimensionality_reduction import dimensionality_reduction
+
+
+
         def print_status(message):
             if verbose:
                 print(message)
@@ -929,6 +943,9 @@ class CustomDataFrame(pd.DataFrame):
         import pandas as pd
         import numpy as np
 
+
+        from sovai.extensions.weight_optimization import WeightOptimization
+
         # Check if self is a pandas DataFrame
         if not isinstance(self, pd.DataFrame):
             raise TypeError("Input must be a pandas DataFrame.")
@@ -974,6 +991,8 @@ class CustomDataFrame(pd.DataFrame):
         import pandas as pd
         import numpy as np
 
+        from sovai.extensions.signal_evaluation import SignalEvaluator
+
         # Check if self.df_factor is a pandas DataFrame
         if not isinstance(self, pd.DataFrame):
             raise TypeError("Input must be a pandas DataFrame.")
@@ -1024,6 +1043,9 @@ class CustomDataFrame(pd.DataFrame):
         return CustomDataFrame(avg_shap_values)
     
     def technical_indicators(self):
+
+        from sovai.extensions.technical_indicators import techn_indicators
+
         """
         Computes feature importance using SHAP values based on multiple simulations.
 
@@ -1049,10 +1071,12 @@ class CustomDataFrame(pd.DataFrame):
 import polars as pl
 import pandas as pd
 import numpy as np
-from numpy.linalg import norm
 
 
 def distance_pl(pandas_df, on="ticker"):
+
+    from numpy.linalg import norm
+
     pl_df = pl.from_pandas(
         pandas_df.reset_index()
         if isinstance(pandas_df.index, pd.MultiIndex)

{sovai-0.2.60 → sovai-0.2.62}/sovai/extensions/time_decomposition.py

@@ -3,7 +3,7 @@ import pandas as pd
 import plotly.graph_objects as go
 from plotly.subplots import make_subplots
 from scipy import stats
-from statsforecast.models import MSTL
+
 import dash
 from dash import dcc, html
 from dash.dependencies import Input, Output
@@ -31,6 +31,8 @@ def categorize_trend(
 
 
 def perform_comprehensive_analysis(df_accounting, ticker, feature):
+    from statsforecast.models import MSTL
+    
     df_signal = (
         df_accounting.query(f"ticker == '{ticker}'")
         .reset_index()

{sovai-0.2.60 → sovai-0.2.62}/sovai/reports/news/news_econometric_analysis.py

@@ -6,8 +6,6 @@ import logging
 import ipywidgets as widgets
 import pandas as pd
 import numpy as np
-from statsmodels.tsa.stattools import adfuller, coint, grangercausalitytests
-from statsmodels.tsa.vector_ar.vecm import VECM
 from IPython.display import display, HTML
 from scipy import stats
 import warnings
@@ -48,6 +46,8 @@ def prepare_data(ticker, variables):
 
 
 def adf_test(series):
+    from statsmodels.tsa.stattools import adfuller
+
     """Perform Augmented Dickey-Fuller test for stationarity."""
     result = adfuller(series)
     return pd.Series({
@@ -59,6 +59,9 @@ def adf_test(series):
     })
 
 def cointegration_test(series1, series2):
+
+    from statsmodels.tsa.stattools import coint
+
     """Perform cointegration test between two series."""
     _, p_value, _ = coint(series1, series2)
     return p_value
@@ -100,6 +103,8 @@ def interpret_vecm_coefficient(coef, p_value):
     return f"{strength.capitalize()} {direction} impact"
 
 def granger_causality(data, variables, max_lags=5):
+    from statsmodels.tsa.stattools import  grangercausalitytests
+
     results = {}
     for v1 in variables:
         for v2 in variables:
@@ -189,6 +194,8 @@ def generate_dynamic_report(df_merged, ticker, variables):
     report.append("<p>VECM combines short-term dynamics with long-term equilibrium relationships. It's particularly useful when variables are cointegrated, as it can model both short-term adjustments and long-term convergence to equilibrium.</p>")
     
     try:
+        from statsmodels.tsa.vector_ar.vecm import VECM
+
         max_lags = 10  # Increase the number of lags considered
         logger.info(f"Fitting VECM model with {max_lags} lags")
         model = VECM(df_merged[all_variables], deterministic='co', k_ar_diff=max_lags)

sovai-0.2.62/sovai/token_auth.py

@@ -0,0 +1,25 @@
+from sovai.api_config import ApiConfig, save_key
+# Removed verify_token import from top level
+from sovai.errors.sovai_errors import InvalidCredentialsError
+
+
+def token_auth(token: str):
+    """Authenticates using a token, saves it, and verifies it."""
+
+    # Lazy load verify_token here, just before use
+    from sovai.utils.client_side import verify_token
+
+    ApiConfig.token = token
+    ApiConfig.token_type = "Bearer"
+    save_key() # Assumes save_key doesn't trigger heavy imports
+
+    # Now call the imported function
+    is_valid, user_id = verify_token(verbose=False) # Uses lazy-loaded verify_token
+
+    if not is_valid:
+        # print('invalid') # Keep commented or use logging
+        raise InvalidCredentialsError("Invalid or expired token. Please authenticate.")
+
+    # Optionally, you could return something useful, like the validity or user_id
+    # print(f"Token verified successfully for user ID: {user_id}") # Or use logging
+    # return is_valid, user_id

{sovai-0.2.60 → sovai-0.2.62}/sovai/utils/client_side.py

@@ -7,13 +7,7 @@ from sovai.errors.sovai_errors import InvalidCredentialsError
 import requests
 from sovai.api_config import ApiConfig
 
-# Try to import CustomDataFrame, use regular DataFrame if not available
-try:
-    from sovai.extensions.pandas_extensions import CustomDataFrame
-    HAS_CUSTOM_DATAFRAME = True
-except ImportError:
-    HAS_CUSTOM_DATAFRAME = False
-    CustomDataFrame = pd.DataFrame  # Fallback to regular DataFrame
+
 
 
 ## might do the same thing as token_auth, you might have recreated it.
@@ -224,15 +218,27 @@ def client_side_frame(endpoint, tickers, columns, start_date, end_date):
     name_file = config["name"]
     partitioned = config["partitioned"]
 
-    # Load data
-    df_percentiles = load_data_for_tickers(
+    # Load data - receives a standard pandas DataFrame
+    df_result = load_data_for_tickers(
         base_url, tickers, columns, start_date, end_date, name_file, partitioned
     )
-    
-    if HAS_CUSTOM_DATAFRAME:
-        return CustomDataFrame(df_percentiles)
-    else:
-        return df_percentiles  # Returns a regular pandas DataFrame if CustomDataFrame is not available
+
+    # --- Lazy Load CustomDataFrame Wrapper ---
+    # Try to import and wrap only if the import succeeds
+    try:
+        from sovai.extensions.pandas_extensions import CustomDataFrame
+        # If import successful, wrap the result
+        print("Wrapping result in CustomDataFrame.") # Optional: for debugging
+        return CustomDataFrame(df_result)
+    except ImportError:
+        # If import fails (lean install), return the standard DataFrame
+        print("CustomDataFrame not available (lean install?). Returning standard pandas DataFrame.") # Optional: for debugging
+        return df_result
+    except Exception as e:
+        # Catch other potential errors during wrapping
+        print(f"Error wrapping result in CustomDataFrame: {e}. Returning standard pandas DataFrame.")
+        return df_result
+
     
 
 # def client_side_frame(endpoint, tickers, columns, start_date, end_date):

sovai-0.2.60/sovai/token_auth.py

@@ -1,15 +0,0 @@
-from sovai.api_config import ApiConfig, save_key
-from sovai.utils.client_side import verify_token
-from sovai.errors.sovai_errors import InvalidCredentialsError
-
-
-def token_auth(token: str):
-    ApiConfig.token = token
-    ApiConfig.token_type = "Bearer"
-    save_key()
-    is_valid, user_id = verify_token(verbose=False)
-
-    if not is_valid:
-        # print('invalid')
-        raise InvalidCredentialsError("Invalid or expired token. Please authenticate.")
-