panelbeater 0.0.2__tar.gz → 0.0.3__tar.gz

{panelbeater-0.0.2/panelbeater.egg-info → panelbeater-0.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: panelbeater
-Version: 0.0.2
+Version: 0.0.3
 Summary: A CLI for finding mispriced options.
 Home-page: https://github.com/8W9aG/panelbeater
 Author: Will Sackfield

{panelbeater-0.0.2 → panelbeater-0.0.3}/panelbeater/__init__.py

@@ -1,3 +1,3 @@
 """panelbeater initialisation."""
 
-__VERSION__ = "0.0.2"
+__VERSION__ = "0.0.3"

{panelbeater-0.0.2 → panelbeater-0.0.3}/panelbeater/__main__.py

@@ -3,11 +3,12 @@
 import datetime
 
 import requests_cache
+import tqdm
 import wavetrainer as wt
 
 from .download import download
 from .features import features
-from .normalizer import normalize
+from .normalizer import denormalize, normalize
 
 _TICKERS = [
     # Equities
@@ -46,6 +47,7 @@ _WINDOWS = [
     200,
 ]
 _LAGS = [1, 3, 5, 10, 20, 30]
+_DAYS_OUT = 30
 
 
 def main() -> None:
@@ -61,8 +63,13 @@ def main() -> None:
     )
     df_y = download(tickers=_TICKERS, macros=_MACROS, session=session)
     df_x = features(df=df_y.copy(), windows=_WINDOWS, lags=_LAGS)
-    df_y = normalize(df=df_y)
-    wavetrainer.fit(df_x, y=df_y)
+    df_y_norm = normalize(df=df_y.copy())
+    for _ in tqdm.tqdm(range(_DAYS_OUT), desc="Running t+X simulation"):
+        wavetrainer.fit(df_x, y=df_y_norm)
+        df_next = wavetrainer.transform(df_y_norm)
+        df_next = denormalize(df_next)
+        df_x = features(df=df_next.copy(), windows=_WINDOWS, lags=_LAGS)
+        df_y_norm = normalize(df=df_next.copy())
 
 
 if __name__ == "__main__":

{panelbeater-0.0.2 → panelbeater-0.0.3}/panelbeater/download.py

@@ -63,8 +63,4 @@ def download(
     levels = pd.concat(
         [prices.add_prefix("PX_"), macro.add_prefix("MACRO_")], axis=1
     ).ffill()
-    return (
-        levels.replace([np.inf, -np.inf], np.nan)
-        .pct_change(fill_method=None)
-        .replace([np.inf, -np.inf], np.nan)
-    )
+    return levels.replace([np.inf, -np.inf], np.nan)

{panelbeater-0.0.2 → panelbeater-0.0.3}/panelbeater/features.py

@@ -36,7 +36,6 @@ def _meta_ticker_feature(
     for window in tqdm.tqdm(windows, desc="Generating window features"):
         dfs.append(df.rolling(window).mean().add_suffix(f"_rmean{window}"))  # type: ignore
         dfs.append(df.rolling(window).std().add_suffix(f"_rstd{window}"))  # type: ignore
-    dfs.append(df.diff().add_suffix("_diff1"))
     return pd.concat(dfs, axis=1).replace([np.inf, -np.inf], np.nan)
 
 

panelbeater-0.0.3/panelbeater/normalizer.py

@@ -0,0 +1,59 @@
+"""Normalize the Y targets to standard deviations."""
+
+# pylint: disable=too-many-locals
+
+import math
+
+import numpy as np
+import pandas as pd
+import tqdm
+from wavetrainer.model.model import PROBABILITY_COLUMN_PREFIX
+
+
+def normalize(df: pd.DataFrame) -> pd.DataFrame:
+    """Normalize the dataframe per column by z-score bucketing."""
+    df = df.pct_change(fill_method=None).replace([np.inf, -np.inf], np.nan)
+    mu = df.rolling(365).mean()
+    sigma = df.rolling(365).std()
+    df = ((((df - mu) / sigma) * 2.0).round() / 2.0).clip(-3, 3)
+    dfs = []
+    for col in tqdm.tqdm(df.columns, desc="Normalising targets"):
+        for unique_val in df[col].unique():
+            if math.isnan(unique_val):
+                continue
+            s = (df[col] == unique_val).rename(f"{col}_{unique_val}")
+            dfs.append(s)
+    return pd.concat(dfs, axis=1)
+
+
+def denormalize(df: pd.DataFrame) -> pd.DataFrame:
+    """Denormalize the dataframe back to a total value."""
+    date_to_add = df.index[-1] + pd.Timedelta(days=1)
+
+    cols = set(df.columns.values.tolist())
+    target_cols = {x for x in cols if "_".join(x.split("_")[:-1])}
+    for col in target_cols:
+        df[col] = None
+
+        # Find the standard deviations
+        z_cols = {x for x in cols if x.startswith(col)}
+        stds = sorted([float(x.replace(col, "").split("_")[1]) for x in z_cols])
+
+        # Find the highest probability standard deviation
+        highest_std_value = 0.0
+        highest_std = None
+        for std in stds:
+            std_suffix = f"{col}_{std}_{PROBABILITY_COLUMN_PREFIX}"
+            std_true_col = sorted([x for x in cols if x.startswith(std_suffix)])[-1]
+            std_value = df[std_true_col].iloc[-1]
+            if std_value > highest_std_value:
+                highest_std_value = std_value
+                highest_std = std
+
+        # Convert the standard deviation back to a value
+        mu = df[col].rolling(365).mean()
+        sigma = df[col].rolling(365).std()
+        value = (highest_std * sigma) + mu
+        df.loc[date_to_add, col] = value
+
+    return df.drop(columns=list(cols))

{panelbeater-0.0.2 → panelbeater-0.0.3/panelbeater.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: panelbeater
-Version: 0.0.2
+Version: 0.0.3
 Summary: A CLI for finding mispriced options.
 Home-page: https://github.com/8W9aG/panelbeater
 Author: Will Sackfield

{panelbeater-0.0.2 → panelbeater-0.0.3}/setup.py

@@ -23,7 +23,7 @@ def install_requires() -> typing.List[str]:
 
 setup(
     name='panelbeater',
-    version='0.0.2',
+    version='0.0.3',
     description='A CLI for finding mispriced options.',
     long_description=long_description,
     long_description_content_type='text/markdown',

panelbeater-0.0.2/panelbeater/normalizer.py

@@ -1,21 +0,0 @@
-"""Normalize the Y targets to standard deviations."""
-
-import math
-
-import pandas as pd
-import tqdm
-
-
-def normalize(df: pd.DataFrame) -> pd.DataFrame:
-    """Normalize the dataframe per column by z-score bucketing."""
-    mu = df.rolling(365).mean()
-    sigma = df.rolling(365).std()
-    df = ((((df - mu) / sigma) * 2.0).round() / 2.0).clip(-3, 3)
-    dfs = []
-    for col in tqdm.tqdm(df.columns, desc="Normalising targets"):
-        for unique_val in df[col].unique():
-            if math.isnan(unique_val):
-                continue
-            s = (df[col] == unique_val).rename(f"{col}_{unique_val}")
-            dfs.append(s)
-    return pd.concat(dfs, axis=1)