Yuxuan Yan
0c524c6987
Joins the bar dataset with one or more alpha parquet files on (symbol, date) for a given symbol and date range. Standalone tools/alphaview.py wired as `cli.py alphaview`, plus README docs. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
59 lines
2.5 KiB
Python
Executable File
59 lines
2.5 KiB
Python
Executable File
#!/usr/bin/env python3
|
|
"""Show alpha weight(s) alongside daily bar data for a single symbol.
|
|
|
|
Joins the bar dataset and one or more alpha parquet files on (symbol, date) so
|
|
you can eyeball how a weight moves against price/volume over a time range.
|
|
"""
|
|
|
|
import click
|
|
import pandas as pd
|
|
|
|
|
|
def _filter_dates(df, start_date, end_date):
|
|
if start_date:
|
|
df = df[df["date"] >= pd.Timestamp(start_date)]
|
|
if end_date:
|
|
df = df[df["date"] <= pd.Timestamp(end_date)]
|
|
return df
|
|
|
|
|
|
@click.command(name="alphaview")
|
|
@click.option("--data-path", required=True, type=click.Path(exists=True), help="Bar dataset dir or parquet file")
|
|
@click.option("--alpha-path", "alpha_paths", required=True, help="Comma-separated alpha parquet path(s)")
|
|
@click.option("--symbol", required=True, help="Symbol id, e.g. sh600000")
|
|
@click.option("--start-date", default=None, help="YYYY-MM-DD (inclusive)")
|
|
@click.option("--end-date", default=None, help="YYYY-MM-DD (inclusive)")
|
|
@click.option("-c", "--columns", default="close,volume", help="Comma-separated bar columns to show")
|
|
def alphaview(data_path, alpha_paths, symbol, start_date, end_date, columns):
|
|
"""Print bar columns + alpha weight(s) for SYMBOL over a date range."""
|
|
bar_cols = [c.strip() for c in columns.split(",") if c.strip()]
|
|
|
|
bars = pd.read_parquet(data_path)
|
|
bars = bars[bars["symbol_id"] == symbol]
|
|
if bars.empty:
|
|
raise click.ClickException(f"Symbol {symbol!r} not found in {data_path}")
|
|
missing = [c for c in bar_cols if c not in bars.columns]
|
|
if missing:
|
|
raise click.ClickException(f"Bar columns not found: {', '.join(missing)}")
|
|
bars = _filter_dates(bars, start_date, end_date)
|
|
out = bars[["date", *bar_cols]].set_index("date").sort_index()
|
|
|
|
for path in (p.strip() for p in alpha_paths.split(",") if p.strip()):
|
|
adf = pd.read_parquet(path)
|
|
adf = adf[adf["symbol_id"] == symbol]
|
|
adf = _filter_dates(adf, start_date, end_date)
|
|
# One alpha file may carry multiple alpha_name values; spread to columns.
|
|
wide = adf.pivot_table(index="date", columns="alpha_name", values="weight")
|
|
out = out.join(wide, how="left")
|
|
|
|
if out.empty:
|
|
raise click.ClickException("No rows in the requested date range")
|
|
|
|
with pd.option_context("display.max_rows", None, "display.max_columns", None, "display.width", None):
|
|
click.echo(f"symbol: {symbol} rows: {len(out):,}")
|
|
click.echo(out.to_string())
|
|
|
|
|
|
if __name__ == "__main__":
|
|
alphaview()
|