Sweep-level deflated Sharpe ratio diagnostic

ledgr_sweep_dsr() computes a native Deflated Sharpe Ratio (DSR) diagnostic over retained completed-candidate return panels. It is an evidence surface only: it does not select, promote, filter, or change walk-forward identity.

Usage

ledgr_sweep_dsr(
  sweep,
  candidates = NULL,
  effective_trials = NULL,
  distance_threshold = 0.5,
  confidence = 0.95,
  risk_free_return = 0
)

# S3 method for class 'ledgr_sweep_dsr'
as_tibble(x, ...)

# S3 method for class 'ledgr_sweep_dsr'
print(x, ...)

Arguments

sweep

A ledgr_sweep_results object with retained completed returns.

candidates

Optional character vector of candidate ids to include.

effective_trials

Optional whole-number effective independent trial count. When NULL, ledgr derives it from ledgr_sweep_cluster().

distance_threshold

Numeric scalar in [0, 2] passed to ledgr_sweep_cluster() when effective_trials is NULL.

confidence

Numeric scalar in (0, 1) used for the significant status flag.

risk_free_return

Numeric scalar per-period risk-free return to subtract before computing Sharpe.

x

A ledgr_sweep_dsr object.

...

Additional arguments passed to tibble printing.

Value

A ledgr_sweep_dsr object with summary and metadata.

Details

The diagnostic consumes ledgr_sweep_returns_panel(..., value = "returns", complete = TRUE). It computes per-period Sharpe, skewness, kurtosis, the variance of candidate Sharpe ratios, and an effective independent trial count. When effective_trials is omitted, the count comes from deterministic retained-return clustering. The output reports a DSR probability and p-value; it does not prove live profitability or automate promotion.

See also

vignette("selection-integrity", package = "ledgr") or system.file("doc", "selection-integrity.html", package = "ledgr").

Examples

if (FALSE) { # \dontrun{
dsr <- ledgr_sweep_dsr(sweep)
as_tibble(dsr)
} # }