Package index • ppidyom

Section

These tools .

build_online_ltm_timestep_counts(): Build Per-Timestep LTM Count Tables with Online Update
compute_local_probs(): Compute Local Probabilities for a Single Order; Used by Backoff PPM
count_tables(): Compute Count Tables for STM and LTM with Optional Prior
lag_matrix(): Generate Lagged N-gram Matrix
ppidyom-class ppidyom: Initialize a new PPM counter Train on a single sequence (incremental LTM update)
apply_prior_ltm(): Seed LTM environments from previously accumulated count tables
envs_to_ltm_tables(): Convert online-update environments to LTM data.table format
precompute_contexts(): Precompute context identifiers for all orders
ltm_to_timestep_counts(): Expand LTM Count Tables to Timestep-Aligned Order Tables
ppm_backoff(): Compute PPM Probabilities for all symbols with Backoff
ppm_interpolated(): Compute PPM Probabilities for all symbols with Interpolation
update_env(): Update a Sparse Context Count Environment