bullishpy 0.13.0__py3-none-any.whl → 0.14.0__py3-none-any.whl

bullish/analysis/filter.py

@@ -1,6 +1,6 @@
 import datetime
 from datetime import date
-from typing import Literal, get_args, Any, Optional, List, Tuple, Type, Dict
+from typing import get_args, Any, Optional, List, Tuple, Type, Dict
 
 from bearish.types import SeriesLength  # type: ignore
 from pydantic import BaseModel, Field, ConfigDict
@@ -14,411 +14,8 @@ from bullish.analysis.analysis import (
     TechnicalAnalysis,
     AnalysisView,
 )
+from bullish.analysis.constants import Industry, IndustryGroup, Sector, Country
 
-Industry = Literal[
-    "Publishing",
-    "Internet Retail",
-    "Scientific & Technical Instruments",
-    "Engineering & Construction",
-    "Diagnostics & Research",
-    "Software - Infrastructure",
-    "Thermal Coal",
-    "Software - Application",
-    "Auto Manufacturers",
-    "Farm Products",
-    "Medical Devices",
-    "Education & Training Services",
-    "Auto Parts",
-    "Specialty Chemicals",
-    "Marine Shipping",
-    "Biotechnology",
-    "Real Estate Services",
-    "Gold",
-    "Entertainment",
-    "Specialty Retail",
-    "Utilities - Independent Power Producers",
-    "Steel",
-    "Mortgage Finance",
-    "Communication Equipment",
-    "Drug Manufacturers - Specialty & Generic",
-    "Electronic Gaming & Multimedia",
-    "Banks - Regional",
-    "Oil & Gas E&P",
-    "Travel Services",
-    "Real Estate - Diversified",
-    "Telecom Services",
-    "Uranium",
-    "Consulting Services",
-    "Waste Management",
-    "Agricultural Inputs",
-    "Utilities - Diversified",
-    "Auto & Truck Dealerships",
-    "Confectioners",
-    "Other Industrial Metals & Mining",
-    "Beverages - Wineries & Distilleries",
-    "Oil & Gas Midstream",
-    "Recreational Vehicles",
-    "Electrical Equipment & Parts",
-    "Household & Personal Products",
-    "Packaging & Containers",
-    "REIT - Specialty",
-    "Home Improvement Retail",
-    "Electronic Components",
-    "Asset Management",
-    "Consumer Electronics",
-    "Conglomerates",
-    "Health Information Services",
-    "Medical Instruments & Supplies",
-    "Building Products & Equipment",
-    "Information Technology Services",
-    "Specialty Industrial Machinery",
-    "Food Distribution",
-    "Packaged Foods",
-    "Rental & Leasing Services",
-    "Medical Distribution",
-    "Grocery Stores",
-    "Advertising Agencies",
-    "Beverages - Non - Alcoholic",
-    "Apparel Manufacturing",
-    "Oil & Gas Equipment & Services",
-    "Coking Coal",
-    "Industrial Distribution",
-    "Restaurants",
-    "Beverages - Brewers",
-    "Chemicals",
-    "Real Estate - Development",
-    "Credit Services",
-    "Tobacco",
-    "Metal Fabrication",
-    "Building Materials",
-    "Residential Construction",
-    "Specialty Business Services",
-    "REIT - Hotel & Motel",
-    "Internet Content & Information",
-    "Lodging",
-    "Furnishings, Fixtures & Appliances",
-    "Airlines",
-    "Computer Hardware",
-    "Integrated Freight & Logistics",
-    "Solar",
-    "Capital Markets",
-    "Leisure",
-    "Airports & Air Services",
-    "Aluminum",
-    "Insurance Brokers",
-    "Semiconductors",
-    "REIT - Retail",
-    "Luxury Goods",
-    "Lumber & Wood Production",
-    "REIT - Mortgage",
-    "Semiconductor Equipment & Materials",
-    "Aerospace & Defense",
-    "Security & Protection Services",
-    "Utilities - Renewable",
-    "Utilities - Regulated Gas",
-    "Apparel Retail",
-    "Pollution & Treatment Controls",
-    "Broadcasting",
-    "Resorts & Casinos",
-    "Other Precious Metals & Mining",
-    "Financial Data & Stock Exchanges",
-    "Footwear & Accessories",
-    "Medical Care Facilities",
-    "Electronics & Computer Distribution",
-    "Gambling",
-    "Tools & Accessories",
-    "Insurance - Property & Casualty",
-    "Utilities - Regulated Water",
-    "Insurance - Specialty",
-    "Personal Services",
-    "Pharmaceutical Retailers",
-    "Farm & Heavy Construction Machinery",
-    "Utilities - Regulated Electric",
-    "Department Stores",
-    "Staffing & Employment Services",
-    "Textile Manufacturing",
-    "Silver",
-    "REIT - Industrial",
-    "REIT - Diversified",
-    "Copper",
-    "Business Equipment & Supplies",
-    "Infrastructure Operations",
-    "Trucking",
-    "Insurance - Reinsurance",
-    "Insurance - Diversified",
-    "Drug Manufacturers - General",
-    "Oil & Gas Drilling",
-    "Banks - Diversified",
-    "REIT - Residential",
-    "Oil & Gas Refining & Marketing",
-    "Shell Companies",
-    "Financial Conglomerates",
-    "Paper & Paper Products",
-    "Insurance - Life",
-    "REIT - Office",
-    "Railroads",
-    "Oil & Gas Integrated",
-    "Healthcare Plans",
-    "REIT - Healthcare Facilities",
-    "Discount Stores",
-]
-
-IndustryGroup = Literal[
-    "publishing",
-    "internet-retail",
-    "scientific-technical-instruments",
-    "engineering-construction",
-    "diagnostics-research",
-    "software-infrastructure",
-    "thermal-coal",
-    "software-application",
-    "auto-manufacturers",
-    "farm-products",
-    "medical-devices",
-    "education-training-services",
-    "auto-parts",
-    "specialty-chemicals",
-    "marine-shipping",
-    "biotechnology",
-    "real-estate-services",
-    "gold",
-    "entertainment",
-    "specialty-retail",
-    "utilities-independent-power-producers",
-    "steel",
-    "mortgage-finance",
-    "communication-equipment",
-    "drug-manufacturers-specialty-generic",
-    "electronic-gaming-multimedia",
-    "banks-regional",
-    "oil-gas-e-p",
-    "travel-services",
-    "real-estate-diversified",
-    "telecom-services",
-    "uranium",
-    "consulting-services",
-    "waste-management",
-    "agricultural-inputs",
-    "utilities-diversified",
-    "auto-truck-dealerships",
-    "confectioners",
-    "other-industrial-metals-mining",
-    "beverages-wineries-distilleries",
-    "oil-gas-midstream",
-    "recreational-vehicles",
-    "electrical-equipment-parts",
-    "household-personal-products",
-    "packaging-containers",
-    "reit-specialty",
-    "home-improvement-retail",
-    "electronic-components",
-    "asset-management",
-    "consumer-electronics",
-    "conglomerates",
-    "health-information-services",
-    "medical-instruments-supplies",
-    "building-products-equipment",
-    "information-technology-services",
-    "specialty-industrial-machinery",
-    "food-distribution",
-    "packaged-foods",
-    "rental-leasing-services",
-    "medical-distribution",
-    "grocery-stores",
-    "advertising-agencies",
-    "beverages-non-alcoholic",
-    "apparel-manufacturing",
-    "oil-gas-equipment-services",
-    "coking-coal",
-    "industrial-distribution",
-    "restaurants",
-    "beverages-brewers",
-    "chemicals",
-    "real-estate-development",
-    "credit-services",
-    "tobacco",
-    "metal-fabrication",
-    "building-materials",
-    "residential-construction",
-    "specialty-business-services",
-    "reit-hotel-motel",
-    "internet-content-information",
-    "lodging",
-    "furnishings-fixtures-appliances",
-    "airlines",
-    "computer-hardware",
-    "integrated-freight-logistics",
-    "solar",
-    "capital-markets",
-    "leisure",
-    "airports-air-services",
-    "aluminum",
-    "insurance-brokers",
-    "semiconductors",
-    "reit-retail",
-    "luxury-goods",
-    "lumber-wood-production",
-    "reit-mortgage",
-    "semiconductor-equipment-materials",
-    "aerospace-defense",
-    "security-protection-services",
-    "utilities-renewable",
-    "utilities-regulated-gas",
-    "apparel-retail",
-    "pollution-treatment-controls",
-    "broadcasting",
-    "resorts-casinos",
-    "other-precious-metals-mining",
-    "financial-data-stock-exchanges",
-    "footwear-accessories",
-    "medical-care-facilities",
-    "electronics-computer-distribution",
-    "gambling",
-    "tools-accessories",
-    "insurance-property-casualty",
-    "utilities-regulated-water",
-    "insurance-specialty",
-    "personal-services",
-    "pharmaceutical-retailers",
-    "farm-heavy-construction-machinery",
-    "utilities-regulated-electric",
-    "department-stores",
-    "staffing-employment-services",
-    "textile-manufacturing",
-    "silver",
-    "reit-industrial",
-    "reit-diversified",
-    "copper",
-    "business-equipment-supplies",
-    "infrastructure-operations",
-    "trucking",
-    "insurance-reinsurance",
-    "insurance-diversified",
-    "drug-manufacturers-general",
-    "oil-gas-drilling",
-    "banks-diversified",
-    "reit-residential",
-    "oil-gas-refining-marketing",
-    "shell-companies",
-    "financial-conglomerates",
-    "paper-paper-products",
-    "insurance-life",
-    "reit-office",
-    "railroads",
-    "oil-gas-integrated",
-    "healthcare-plans",
-    "reit-healthcare-facilities",
-    "discount-stores",
-]
-
-Sector = Literal[
-    "Communication Services",
-    "Consumer Cyclical",
-    "Technology",
-    "Industrials",
-    "Healthcare",
-    "Energy",
-    "Consumer Defensive",
-    "Basic Materials",
-    "Real Estate",
-    "Utilities",
-    "Financial Services",
-    "Conglomerates",
-]
-
-Country = Literal[
-    "Australia",
-    "China",
-    "Japan",
-    "United kingdom",
-    "United states",
-    "Poland",
-    "Switzerland",
-    "Canada",
-    "Greece",
-    "Spain",
-    "Germany",
-    "Indonesia",
-    "Belgium",
-    "France",
-    "Netherlands",
-    "British virgin islands",
-    "Italy",
-    "Hungary",
-    "Austria",
-    "Finland",
-    "Sweden",
-    "Bermuda",
-    "Taiwan",
-    "Israel",
-    "Ukraine",
-    "Singapore",
-    "Jersey",
-    "Ireland",
-    "Luxembourg",
-    "Cyprus",
-    "Cayman islands",
-    "Norway",
-    "Denmark",
-    "Hong kong",
-    "New zealand",
-    "Kazakhstan",
-    "Nigeria",
-    "Argentina",
-    "Brazil",
-    "Czech republic",
-    "Mauritius",
-    "South africa",
-    "India",
-    "Mexico",
-    "Mongolia",
-    "Slovenia",
-    "Thailand",
-    "Malaysia",
-    "Costa rica",
-    "Isle of man",
-    "Egypt",
-    "Turkey",
-    "United arab emirates",
-    "Colombia",
-    "Gibraltar",
-    "Malta",
-    "Liechtenstein",
-    "Guernsey",
-    "Peru",
-    "Estonia",
-    "French guiana",
-    "Portugal",
-    "Uruguay",
-    "Chile",
-    "Martinique",
-    "Monaco",
-    "Panama",
-    "Papua new guinea",
-    "South korea",
-    "Macau",
-    "Gabon",
-    "Romania",
-    "Senegal",
-    "Morocco",
-    "Jordan",
-    "Lithuania",
-    "Dominican republic",
-    "Reunion",
-    "Zambia",
-    "Cambodia",
-    "Myanmar",
-    "Bahamas",
-    "Philippines",
-    "Bangladesh",
-    "Latvia",
-    "Vietnam",
-    "Iceland",
-    "Azerbaijan",
-    "Georgia",
-    "Liberia",
-    "Kenya",
-]
 SIZE_RANGE = 2
 
 

bullish/analysis/functions.py

@@ -1,4 +1,3 @@
-import datetime
 import logging
 from datetime import date
 from typing import Optional, Callable, cast
@@ -282,24 +281,6 @@ def compute_price(data: pd.DataFrame) -> pd.DataFrame:
     return results
 
 
-def compute_percentile_return_after_rsi_crossover(
-    data: pd.DataFrame, rsi_threshold: int = 45, period: int = 90
-) -> float:
-    data_ = cross_value_series(data.RSI, rsi_threshold)
-    values = []
-    for crossing_date in data_[data_ == 1].index:
-        data_crossed = data[
-            (data.index >= crossing_date)
-            & (data.index <= crossing_date + datetime.timedelta(days=period))
-        ]
-        v = (
-            data_crossed.CLOSE.pct_change(periods=len(data_crossed.CLOSE) - 1).iloc[-1]
-            * 100
-        )
-        values.append(v)
-    return float(np.percentile(values, 30))
-
-
 def find_last_true_run_start(series: pd.Series) -> Optional[date]:
     if not series.iloc[-1]:
         return None
@@ -311,6 +292,16 @@ def find_last_true_run_start(series: pd.Series) -> Optional[date]:
     return series.index[last_true_run_start].date()  # type: ignore
 
 
+def sma_50_above_sma_200(data: pd.DataFrame) -> Optional[date]:
+    date_1 = find_last_true_run_start(data.SMA_50 > data.SMA_200)
+    return date_1
+
+
+def price_above_sma50(data: pd.DataFrame) -> Optional[date]:
+    date_1 = find_last_true_run_start(data.SMA_50 < data.CLOSE)
+    return date_1
+
+
 def momentum(data: pd.DataFrame) -> Optional[date]:
     date_1 = find_last_true_run_start(data.SMA_50 < data.CLOSE)
     date_2 = find_last_true_run_start(data.SMA_200 < data.SMA_50)

bullish/analysis/indicators.py

@@ -19,8 +19,9 @@ from bullish.analysis.functions import (
     SMA,
     ADOSC,
     PRICE,
-    compute_percentile_return_after_rsi_crossover,
     momentum,
+    sma_50_above_sma_200,
+    price_above_sma50,
 )
 
 logger = logging.getLogger(__name__)
@@ -206,13 +207,6 @@ def indicators_factory() -> List[Indicator]:
                         (d.RSI < 60) & (d.RSI > 40)
                     ].last_valid_index(),
                 ),
-                Signal(
-                    name="RETURN_AFTER_RSI_CROSSOVER_45_PERIOD_90",
-                    description="Percentile 30 return after RSI crossover 45 in the next 90 days",
-                    type_info="Long",
-                    type=Optional[float],
-                    function=lambda d: compute_percentile_return_after_rsi_crossover(d),
-                ),
             ],
         ),
         Indicator(
@@ -290,6 +284,20 @@ def indicators_factory() -> List[Indicator]:
                     type=Optional[date],
                     function=lambda d: momentum(d),
                 ),
+                Signal(
+                    name="SMA_50_ABOVE_SMA_200",
+                    description="SMA 50 is above SMA 200",
+                    type_info="Overbought",
+                    type=Optional[date],
+                    function=lambda d: sma_50_above_sma_200(d),
+                ),
+                Signal(
+                    name="PRICE_ABOVE_SMA_50",
+                    description="Price is above SMA 50",
+                    type_info="Overbought",
+                    type=Optional[date],
+                    function=lambda d: price_above_sma50(d),
+                ),
             ],
         ),
         Indicator(

bullish/analysis/industry_views.py

@@ -0,0 +1,201 @@
+import datetime
+from typing import (
+    Optional,
+    Any,
+    Annotated,
+    Literal,
+    Dict,
+    List,
+    TYPE_CHECKING,
+    get_args,
+)
+
+import numpy as np
+import pandas as pd
+from bearish.models.base import Ticker  # type: ignore
+from bearish.models.price.prices import Prices  # type: ignore
+from bearish.models.query.query import AssetQuery, Symbols  # type: ignore
+from pydantic import BaseModel, BeforeValidator, Field, model_validator
+
+from bullish.analysis.constants import Industry, IndustryGroup, Sector, Country
+
+if TYPE_CHECKING:
+    from bullish.database.crud import BullishDb
+
+Type = Literal["Mean"]
+
+FUNCTIONS = {"Mean": np.mean}
+BASELINE_DATE = datetime.date.today() - datetime.timedelta(days=60)
+
+
+def compute_normalized_close(close_: pd.Series) -> pd.Series:
+    close = close_.copy()
+    close.index = close.index.tz_localize(None)  # type: ignore
+    closest_ts = close.index[
+        close.index.get_indexer([BASELINE_DATE], method="nearest")[0]
+    ]
+    normalized_close = (close / close.loc[closest_ts]).rename("normalized_close")
+    normalized_close.index = close_.index
+    return normalized_close  # type: ignore
+
+
+def get_industry_comparison_data(
+    bullish_db: "BullishDb",
+    symbol_data: pd.DataFrame,
+    type: Type,
+    industry: Industry,
+    country: Country,
+) -> pd.DataFrame:
+    try:
+        views = bullish_db.read_returns(type, industry, country)
+        industry_data = IndustryViews.from_views(views).to_dataframe()
+        normalized_symbol = compute_normalized_close(symbol_data.close).rename("symbol")
+        normalized_industry = industry_data.normalized_close.rename("industry")
+        return pd.concat([normalized_symbol, normalized_industry], axis=1)
+    except Exception:
+        return pd.DataFrame()
+
+
+class PricesReturns(Prices):  # type: ignore
+
+    def returns(self) -> pd.DataFrame:
+        try:
+            data = self.to_dataframe()
+            data["simple_return"] = data.close.pct_change() * 100
+            data["log_return"] = (data.close / data.close.shift(1)).apply(np.log) * 100
+            data["normalized_close"] = compute_normalized_close(data.close)
+            return data[["simple_return", "log_return", "normalized_close"]]  # type: ignore
+        except Exception:
+            return pd.DataFrame(
+                columns=["simple_return", "log_return", "normalized_close"]
+            )
+
+
+def to_float(value: Any) -> Optional[float]:
+    if value == "None":
+        return None
+    if value is None:
+        return None
+    if isinstance(value, str):
+        try:
+            return float(value)
+        except ValueError:
+            return None
+    return float(value)
+
+
+class Basedate(BaseModel):
+    date: datetime.date
+
+
+class KPI(BaseModel):
+    simple_return: Annotated[float, BeforeValidator(to_float), Field(None)]
+    log_return: Annotated[float, BeforeValidator(to_float), Field(None)]
+    normalized_close: Annotated[float, BeforeValidator(to_float), Field(None)]
+
+
+class BaseIndustryView(Basedate, KPI): ...
+
+
+class IndustryView(BaseIndustryView):
+    created_at: datetime.date
+    country: Country
+    industry: Industry
+    industry_group: Optional[IndustryGroup] = None
+    sector: Optional[Sector] = None
+    type: Type
+
+    @model_validator(mode="before")
+    def _validate(cls, values: Dict[str, Any]) -> Dict[str, Any]:  # noqa: N805
+        created_at = datetime.date.today()
+        current_date = values.get("date", created_at)
+        return (
+            {"date": current_date}
+            | values
+            | {
+                "created_at": created_at,
+            }
+        )
+
+    @classmethod
+    def from_data(
+        cls,
+        data: pd.DataFrame,
+        function_name: Type,
+        industry: Industry,
+        country: Country,
+    ) -> List["IndustryView"]:
+        function = FUNCTIONS[function_name]
+        data_ = []
+        for field in KPI.model_fields:
+
+            data__ = (
+                data[field].apply(function, axis=1).rename(field)
+                if data[[field]].shape[1] > 1
+                else data[field]
+            )
+
+            data_.append(data__)
+
+        data_final = pd.concat(data_, axis=1)
+        data_final["date"] = data_final.index
+        return [
+            cls.model_validate(
+                r | {"industry": industry, "type": function_name, "country": country}
+            )
+            for r in data_final.to_dict(orient="records")
+        ]
+
+    @classmethod
+    def from_db(
+        cls, bullish: "BullishDb", industry: Industry, country: Country
+    ) -> List["IndustryView"]:
+        returns = []
+        symbols = bullish.read_industry_symbols(industries=[industry], country=country)
+        query = AssetQuery(
+            symbols=Symbols(equities=[Ticker(symbol=s) for s in symbols])
+        )
+        data = bullish.read_series(query, months=6)
+        raw_data = [
+            PricesReturns(prices=[d for d in data if d.symbol == s]).returns()
+            for s in symbols
+        ]
+        raw_data = [r for r in raw_data if not r.empty]
+
+        if raw_data:
+            data_ = pd.concat(raw_data, axis=1)
+            for function_name in FUNCTIONS:
+                returns.extend(cls.from_data(data_, function_name, industry, country))  # type: ignore
+        return returns
+
+
+class IndustryViews(BaseModel):
+    views: List[IndustryView]
+
+    def to_dataframe(self) -> pd.DataFrame:
+        data = pd.DataFrame.from_records(
+            [
+                p.model_dump(include=set(BaseIndustryView.model_fields))
+                for p in self.views
+            ]
+        )
+        if data.empty:
+            return data
+        data = data.set_index("date", inplace=False)
+        data = data.sort_index(inplace=False)
+
+        data.index = pd.to_datetime(data.index, utc=True)
+        data = data[~data.index.duplicated(keep="first")]
+        return data
+
+    @classmethod
+    def from_views(cls, views: List[IndustryView]) -> "IndustryViews":
+        return cls(views=views)
+
+
+def compute_industry_view(bullish: "BullishDb") -> None:
+    for country in get_args(Country):
+        for industry in get_args(Industry):
+            returns = IndustryView.from_db(bullish, industry, country)
+            if returns:
+                bullish.write_returns(returns)