API overview

Literature mapping lives in Methods; DOIs in References. Docstrings on the symbols below also carry # References sections.

Small-n profile

recommend_folds · recommend_run_options · recommend_learners
SMALL_N_SL_LEARNERS · adaptive_learners · warn_if_folds_too_large

LMTP and policies

run_lmtp_grid · lmtp_tmle_contrast · ShiftPolicy
additive_shift_policy · multiplicative_shift_policy · threshold_shift_policy
SequentialPolicy · run_sequential_lmtp · sequential_identification_certificate
build_lmtp_fold_cache

Mediation

run_mediation_grid · run_mediation_scalar
mediation_n_mc_sweep · mediation_stability_summary · mediation_stability_markdown
build_mediation_fold_cache · MediationFoldCache
Soft-deprecated aliases (emit DeprecationWarning): run_crumble_grid, run_crumble_scalar, build_crumble_fold_cache, CrumbleFoldCache, run_crumble_scalar_ppl
Prefer: run_mediation_grid · run_mediation_scalar · run_mediation_scalar_ppl · MediationFoldCache

Positivity and sensitivity

positivity_report · positivity_markdown · attach_positivity_summary!
tipping_point_bias · partial_r2_calibration · sensitivity_report · sensitivity_markdown
adjustment_set_disagreement · discovery_adjustment_sensitivity · merge_discovery_sensitivity!

Planning and certificates

plan_mtp · summarise_plan · execute_estimand
identification_certificate · certificate_dict
build_run_metadata · attach_run_metadata!

CausalTargeted.CausalTargeted — Module

CausalTargeted

Cross-fitted targeted inference: LMTP, interventional mediation EIF, nuisance caching, and grid execution. Identification is delegated to CausalDynamics.jl.

Small-n profiles, positivity atlases, and sensitivity helpers target conservation and other low-sample causal applications.

Documentation

Methods ↔ literature: docs/src/methods.md
Full bibliography (DOIs / BibTeX keys): docs/src/references.md
Small-n checklist: docs/src/small_n.md

Canonical papers: Díaz et al. (2023) LMTP; Díaz & Hejazi (2020) / Liu et al. (2024) mediation; van der Laan & Rose (2011) TMLE; Cinelli & Hazlett (2020) sensitivity.

source

CausalTargeted.recommend_run_options — Function

recommend_run_options(n; engine, n_mediators, rich) -> NamedTuple

Kwargs suitable for run_lmtp_grid / run_mediation_grid / execute_estimand under a small-n-first policy.

n	folds	learners	density_ratio	n_mc (mediation)	parallel
< 40	2	small	gaussian	64–128	false
40–79	3	default	gaussian	64	false
≥ 80	3–5	default/rich	hybrid if rich	32–64	false*

* Parallel remains off by default for memory safety; set parallel=true explicitly.

source

CausalTargeted.run_lmtp_grid — Function

run_lmtp_grid(data, trt, outcome; baseline, kwargs...) -> DataFrame

Cross-fitted SuperLearner LMTP TMLE grid matching R run_lmtp_grid() (shift_scale = "z").

Uses clamp-aware hybrid targeting: when many observations hit clamp bounds, the TMLE fluctuation is down-weighted toward g-computation.

Defaults:

density_ratio = :gaussian (stable at sheep n; use :hybrid / :classification for large n)
cv_trunc = true (select hard truncation among candidates)
estimator = :tmle (score-solving; pass :sdr / :eif / :itmle as needed)
epochs = 1 (do not inherit mediation-grid epochs=30)
simultaneous = true adds multiplier-bootstrap simultaneous bands
parallel = true when Threads.nthreads() > 1
cache_nuisances = true reuses fold outcome / exposure models across δ

source

CausalTargeted.run_mediation_grid — Function

run_mediation_grid(args...; kwargs...)

Deprecated façade — prefer CausalMediation.run_mediation_grid.

source

CausalTargeted.positivity_report — Function

positivity_report(data, trt; deltas, stratify_by, lower_q, upper_q, shift_scale, kwargs...) -> DataFrame

Tidy atlas of support / clamp diagnostics across the δ-grid and strata.

Uses support_diagnostics and additive_clamp_diagnostics. Statuses: ok, weak_support, unsupported_shift.

source

CausalTargeted.sensitivity_report — Function

sensitivity_report(est, se; n, alpha, r2_grid) -> DataFrame

One-row tipping point plus a small grid of partial-R² calibrations.

source

CausalTargeted.SequentialPolicy — Type

SequentialPolicy(treatments, outcome, baseline, time_vary, shift)

Time-ordered treatments A_1,…,A_T, baseline covariates L_0, and optional time-varying covariates time_vary[t] observed before A_t.

source