factor_tearsheet¶

Defined in fynance.plot

factor_tearsheet(factor, prices, n_quantiles=5, horizons=(1, 5, 10, 21), w=63, figsize=None, save=None)[source]

Build a full Alphalens-style factor tear-sheet figure.

Composes four panels into a 2x2 Figure from a factor panel and its price panel: quantile cumulative returns, the long-short spread equity, the rolling Information Coefficient and the IC decay by horizon. The one-bar forward return (built from prices) is the target for the quantile and rolling-IC panels; the decay panel uses non-overlapping horizons returns.

Alignment. factor[t] is the score known at bar t; the forward returns are built from prices so no future information leaks in.

Parameters:

factorarray-like: Factor panel (T, N).
pricesarray-like: Price panel (T, N) with the same time index as factor.
n_quantilesint, optional: Number of factor quantiles (default 5).
horizonstuple of int, optional: Forward horizons for the IC-decay panel (default (1, 5, 10, 21)).
wint, optional: Trailing window of the rolling IC (default 63).
figsizetuple of float, optional: Figure size; defaults to (11, 7).
savestr, optional: If given, the figure is written to this path with savefig.

Returns:

matplotlib.figure.Figure