tune_ml_backtest: Quick grid tuning for tabular pipeline

In PortfolioTesteR: Test Investment Strategies with English-Like Code

Quick grid tuning for tabular pipeline

Description

Quick grid tuning for tabular pipeline

Usage

tune_ml_backtest(
  features_list,
  labels,
  prices,
  fit_fn,
  predict_fn,
  schedule = list(is = 104L, oos = 4L, step = 4L),
  grid = list(top_k = c(10L, 15L), temperature = c(8, 12), method = c("softmax", "rank"),
    transform = c("zscore")),
  group = "pooled",
  selection_defaults = list(top_k = 15L, max_per_group = NULL),
  weighting_defaults = list(method = "softmax", temperature = 12, floor = 0),
  caps = list(max_per_symbol = 0.08),
  group_map = NULL,
  cost_bps = 0,
  freq = 52
)

Arguments

features_list	List of feature panels.
labels	Label panel.
prices	Price panel used for backtests (Date + symbols).
fit_fn, predict_fn	Model fit and predict functions.
schedule	List with elements is, oos, step.
grid	list of vectors: top_k, temperature, method, transform.
group	Grouping mode for roll_fit_predict ('pooled'/'per_symbol'/'per_group').
selection_defaults	Default selection settings (e.g., top_k).
weighting_defaults	Default weighting settings (e.g., method, temperature).
caps	Exposure caps (e.g., max_per_symbol/max_per_group).
group_map	Optional Symbol->Group mapping.
cost_bps	optional one-way cost in basis points for net performance.
freq	re-annualization frequency (e.g., 52).

Value

data.table with metrics per grid row.

PortfolioTesteR documentation built on Nov. 5, 2025, 5:23 p.m.