Business-objective criteria — ledgr_objective

These constructors create ledgr-owned, serializable criterion steps for ledgr_business_objective(). Each step thresholds already-computed evidence and returns eligibility evidence only. Steps do not rank, select, promote, persist, or alter execution or walk-forward identity.

Usage

ledgr_objective_even_trades(max_concentration = 0.3)

ledgr_objective_even_profit(max_concentration = 0.4)

ledgr_objective_stable_region(
  min_neighbors = 2L,
  metric = "sharpe_ratio",
  direction = c("maximize", "minimize"),
  parameter_columns = NULL
)

ledgr_objective_max_drawdown(max_drawdown = 0.2)

ledgr_objective_stable_runs(max_streak = 10L)

ledgr_objective_min_trades(n = 30L)

ledgr_objective_positive_trajectory(slope_min = 0)

ledgr_objective_diagnostic_threshold(x, column, threshold, comparison = NULL)

Arguments

max_concentration: Finite scalar in (0, 1] giving the largest acceptable concentration share.
min_neighbors: Positive whole-number minimum count of supporting Manhattan-distance-1 lattice neighbors.
metric: Non-empty sweep-summary metric column used by the strict lattice detector.
direction: Either "maximize" or "minimize"; values are normalized so larger detector scores are better.
parameter_columns: Optional non-empty character vector naming the candidate parameter axes. NULL uses all resolved varying parameter columns supplied to the detector.
max_drawdown: Finite scalar in [0, 1] giving the largest acceptable retained maximum drawdown.
max_streak: Positive whole number giving the largest acceptable consecutive winning or losing run.
n: Non-negative whole-number minimum closed-trade count.
slope_min: Finite minimum slope in log-equity units per retained observation.
x: A ledgr_dsr or ledgr_min_track_record result.
column: An admissible candidate-level diagnostic output column.
threshold: A finite numeric threshold for numeric columns, or one required status string for a status column.
comparison: Optional comparison. Numeric defaults are "at_least" for dsr_probability and "at_most" for min_track_record_length; status columns require equality.

Value

A classed internal-contract ledgr_objective_criterion object.

Details

ledgr_objective_even_trades() divides the sweep scoring interval into four equal-duration bins and measures the largest bin's share of closed trades. ledgr_objective_even_profit() measures the largest single trade's share of total absolute closed-trade realized P&L. ledgr_objective_stable_runs() measures the longest consecutive WIN or LOSS run in trade_seq order; BREAKEVEN rows reset a run.

ledgr_objective_stable_region() implements ledgr's strict-lattice operationalization of Pardo's broad-plateau idea. It is not a transcription of a Pardo formula and does not support sparse or unordered parameter spaces. Lattice-boundary candidates have fewer available neighbors and cannot pass when min_neighbors exceeds that available count; the detector retains the available-neighbor count as audit evidence.

ledgr_objective_max_drawdown() accepts ledgr's canonical signed drawdown metric and exposes its positive loss magnitude as the measured criterion value before applying max_drawdown.

ledgr_objective_diagnostic_threshold() embeds a hashed snapshot of one already-computed ledgr_dsr() or ledgr_min_track_record() result. It never recomputes the diagnostic. Admissible columns are dsr_probability and status for DSR, and min_track_record_length and status for MinTRL. Numeric diagnostic thresholds treat Inf and -Inf as determinate boundary evidence; missing values and NaN fail closed.

References

Pardo, R. (2008). The Evaluation and Optimization of Trading Strategies, Chapter 11. Stable-region adjacency and tolerance are ledgr's documented operationalization, not a Pardo-authored formula.

Examples

ledgr_objective_even_trades(0.30)
#> ledgr objective criterion
#> Criterion: even_trades
#> Evidence:  closed_trades
#> Hash:      19a5b751c70d
ledgr_objective_even_profit(0.40)
#> ledgr objective criterion
#> Criterion: even_profit
#> Evidence:  closed_trades
#> Hash:      32eb244c8a09
ledgr_objective_stable_region(min_neighbors = 2)
#> ledgr objective criterion
#> Criterion: stable_region
#> Evidence:  candidate_metric_lattice
#> Hash:      7752d7a76c9b
ledgr_objective_max_drawdown(0.20)
#> ledgr objective criterion
#> Criterion: max_drawdown
#> Evidence:  summary.max_drawdown
#> Hash:      5e4779b48e89
ledgr_objective_stable_runs(10)
#> ledgr objective criterion
#> Criterion: stable_runs
#> Evidence:  closed_trades
#> Hash:      a92ddb2477a5
ledgr_objective_min_trades(30)
#> ledgr objective criterion
#> Criterion: min_trades
#> Evidence:  summary.n_trades
#> Hash:      cba89b76721e
ledgr_objective_positive_trajectory(0)
#> ledgr objective criterion
#> Criterion: positive_trajectory
#> Evidence:  retained_equity
#> Hash:      69d76f8162e1