File size: 10,553 Bytes
a2afe2f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 |
"""Quantitative Analysis Router."""
from typing import Literal
from openbb_core.app.model.example import APIEx, PythonEx
from openbb_core.app.model.obbject import OBBject
from openbb_core.app.router import Router
from openbb_core.provider.abstract.data import Data
from openbb_quantitative.models import (
ADFTestModel,
CAPMModel,
KPSSTestModel,
NormalityModel,
SummaryModel,
TestModel,
UnitRootModel,
)
from openbb_quantitative.performance.performance_router import (
router as performance_router,
)
from openbb_quantitative.rolling.rolling_router import router as rolling_router
from openbb_quantitative.stats.stats_router import router as stats_router
router = Router(prefix="", description="Quantitative analysis tools.")
router.include_router(rolling_router)
router.include_router(stats_router)
router.include_router(performance_router)
@router.command(
methods=["POST"],
examples=[
PythonEx(
description="Get Normality Statistics.",
code=[
"stock_data = obb.equity.price.historical(symbol='TSLA', start_date='2023-01-01', provider='fmp').to_df()", # noqa: E501
"obb.quantitative.normality(data=stock_data, target='close')",
],
),
APIEx(parameters={"target": "close", "data": APIEx.mock_data("timeseries", 8)}),
],
)
def normality(data: list[Data], target: str) -> OBBject[NormalityModel]:
"""Get Normality Statistics.
- **Kurtosis**: whether the kurtosis of a sample differs from the normal distribution.
- **Skewness**: whether the skewness of a sample differs from the normal distribution.
- **Jarque-Bera**: whether the sample data has the skewness and kurtosis matching a normal distribution.
- **Shapiro-Wilk**: whether a random sample comes from a normal distribution.
- **Kolmogorov-Smirnov**: whether two underlying one-dimensional probability distributions differ.
Parameters
----------
data : list[Data]
Time series data.
target : str
Target column name.
Returns
-------
OBBject[NormalityModel]
Normality tests summary. See qa_models.NormalityModel for details.
"""
# pylint: disable=import-outside-toplevel
from scipy import stats # noqa
from openbb_core.app.utils import ( # noqa
basemodel_to_df,
get_target_column,
)
df = basemodel_to_df(data)
series_target = get_target_column(df, target)
kt_statistic, kt_pvalue = stats.kurtosistest(series_target)
sk_statistic, sk_pvalue = stats.skewtest(series_target)
jb_statistic, jb_pvalue = stats.jarque_bera(series_target)
sh_statistic, sh_pvalue = stats.shapiro(series_target)
ks_statistic, ks_pvalue = stats.kstest(series_target, "norm")
norm_summary = NormalityModel(
kurtosis=TestModel(statistic=kt_statistic, p_value=kt_pvalue),
skewness=TestModel(statistic=sk_statistic, p_value=sk_pvalue),
jarque_bera=TestModel(statistic=jb_statistic, p_value=jb_pvalue),
shapiro_wilk=TestModel(statistic=sh_statistic, p_value=sh_pvalue),
kolmogorov_smirnov=TestModel(statistic=ks_statistic, p_value=ks_pvalue),
)
return OBBject(results=norm_summary)
@router.command(
methods=["POST"],
examples=[
PythonEx(
description="Get Capital Asset Pricing Model (CAPM).",
code=[
"stock_data = obb.equity.price.historical(symbol='TSLA', start_date='2023-01-01', provider='fmp').to_df()", # noqa: E501
"obb.quantitative.capm(data=stock_data, target='close')",
],
),
APIEx(
parameters={"target": "close", "data": APIEx.mock_data("timeseries", 31)}
),
],
)
def capm(data: list[Data], target: str) -> OBBject[CAPMModel]:
"""Get Capital Asset Pricing Model (CAPM).
CAPM offers a streamlined way to assess the expected return on an investment while accounting for its risk relative
to the market. It's a cornerstone of modern financial theory that helps investors understand the trade-off between
risk and return, guiding more informed investment choices.
Parameters
----------
data : list[Data]
Time series data.
target : str
Target column name.
Returns
-------
OBBject[CAPMModel]
CAPM model summary.
"""
# pylint: disable=import-outside-toplevel
import statsmodels.api as sm # noqa
from openbb_core.app.utils import ( # noqa``
basemodel_to_df,
get_target_columns,
)
from pandas import to_datetime # noqa
from openbb_quantitative.helpers import get_fama_raw # noqa
df = basemodel_to_df(data)
df_target = get_target_columns(df, ["date", target])
df_target = df_target.set_index("date")
df_target.loc[:, "return"] = df_target.pct_change()
df_target = df_target.dropna()
df_target.index = to_datetime(df_target.index)
start_date = df_target.index.min().strftime("%Y-%m-%d")
end_date = df_target.index.max().strftime("%Y-%m-%d")
df_fama = get_fama_raw(start_date, end_date)
df_target = df_target.merge(df_fama, left_index=True, right_index=True)
df_target["excess_return"] = df_target["return"] - df_target["RF"]
df_target["excess_mkt"] = df_target["MKT-RF"] - df_target["RF"]
df_target = df_target.dropna()
y = df_target[["excess_return"]]
x = df_target["excess_mkt"]
x = sm.add_constant(x)
model = sm.OLS(y, x).fit()
results = CAPMModel(
market_risk=model.params["excess_mkt"],
systematic_risk=model.rsquared,
idiosyncratic_risk=1 - model.rsquared,
)
return OBBject(results=results)
@router.command(
methods=["POST"],
examples=[
PythonEx(
description="Get Unit Root Test.",
code=[
"stock_data = obb.equity.price.historical(symbol='TSLA', start_date='2023-01-01', provider='fmp').to_df()", # noqa: E501
"obb.quantitative.unitroot_test(data=stock_data, target='close')",
],
),
APIEx(parameters={"target": "close", "data": APIEx.mock_data("timeseries", 5)}),
],
)
def unitroot_test(
data: list[Data],
target: str,
fuller_reg: Literal["c", "ct", "ctt", "nc", "c"] = "c",
kpss_reg: Literal["c", "ct"] = "c",
) -> OBBject[UnitRootModel]:
"""Get Unit Root Test.
This function applies two renowned tests to assess whether your data series is stationary or if it contains a unit
root, indicating it may be influenced by time-based trends or seasonality. The Augmented Dickey-Fuller (ADF) test
helps identify the presence of a unit root, suggesting that the series could be non-stationary and potentially
unpredictable over time. On the other hand, the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) test checks for the
stationarity of the series, where failing to reject the null hypothesis indicates a stable, stationary series.
Together, these tests provide a comprehensive view of your data's time series properties, essential for
accurate modeling and forecasting.
Parameters
----------
data : list[Data]
Time series data.
target : str
Target column name.
fuller_reg : Literal["c", "ct", "ctt", "nc", "c"]
Regression type for ADF test.
kpss_reg : Literal["c", "ct"]
Regression type for KPSS test.
Returns
-------
OBBject[UnitRootModel]
Unit root tests summary.
"""
# pylint: disable=import-outside-toplevel
from openbb_core.app.utils import ( # noqa
basemodel_to_df,
get_target_column,
)
from statsmodels.tsa import stattools # noqa
df = basemodel_to_df(data)
series_target = get_target_column(df, target)
adf = stattools.adfuller(series_target, regression=fuller_reg)
kpss = stattools.kpss(series_target, regression=kpss_reg, nlags="auto")
unitroot_summary = UnitRootModel(
adf=ADFTestModel(
statistic=adf[0],
p_value=adf[1],
nlags=adf[2] if isinstance(adf[2], int) else 0,
nobs=adf[3] if isinstance(adf[3], int) else 0,
icbest=adf[5] if isinstance(adf[5], float) else 0.0, # type: ignore
),
kpss=KPSSTestModel(
statistic=kpss[0],
p_value=kpss[1],
nlags=kpss[2],
),
)
return OBBject(results=unitroot_summary)
@router.command(
methods=["POST"],
examples=[
PythonEx(
description="Get Summary Statistics.",
code=[
"stock_data = obb.equity.price.historical(symbol='TSLA', start_date='2023-01-01', provider='fmp').to_df()", # noqa: E501
"obb.quantitative.summary(data=stock_data, target='close')",
],
),
APIEx(parameters={"target": "close", "data": APIEx.mock_data("timeseries", 5)}),
],
)
def summary(data: list[Data], target: str) -> OBBject[SummaryModel]:
"""Get Summary Statistics.
The summary that offers a snapshot of its central tendencies, variability, and distribution.
This command calculates essential statistics, including mean, standard deviation, variance,
and specific percentiles, to provide a detailed profile of your target column. B
y examining these metrics, you gain insights into the data's overall behavior, helping to identify patterns,
outliers, or anomalies. The summary table is an invaluable tool for initial data exploration,
ensuring you have a solid foundation for further analysis or reporting.
Parameters
----------
data : list[Data]
Time series data.
target : str
Target column name.
Returns
-------
OBBject[SummaryModel]
Summary table.
"""
# pylint: disable=import-outside-toplevel
from openbb_core.app.utils import (
basemodel_to_df,
get_target_column,
)
df = basemodel_to_df(data)
series_target = get_target_column(df, target)
df_stats = series_target.describe(percentiles=[0.1, 0.25, 0.5, 0.75, 0.9])
df_stats.loc["var"] = df_stats.loc["std"] ** 2
results = SummaryModel(
count=df_stats.loc["count"],
mean=df_stats.loc["mean"],
std=df_stats.loc["std"],
var=df_stats.loc["var"],
min=df_stats.loc["min"],
p_25=df_stats.loc["25%"],
p_50=df_stats.loc["50%"],
p_75=df_stats.loc["75%"],
max=df_stats.loc["max"],
)
return OBBject(results=results)
|