Skip to contents

Conversation membership and per-modality window spans (Data View v2)

tma.conversations2.Rd

TMA-aware successor to tma.conversations(). For each unit's coded reference rows it returns the co-occurring rows within that row's per-modality window, plus a rowWindows payload that drives the webtool Data View window-span hover.

Usage

tma.conversations2(x, codes = NULL, id_col = "QEID")

Arguments

x

An accumulated tma set (from accumulate()); must carry _function.params$tensor and _function.params$mode_column (populated by accumulate() since the 5a change).

codes

Code column names. Defaults to _function.params$codes.

id_col

Row-identity column. Default "QEID".

Value

A list with unitRows, allRows, toRemove (row ids) and rowWindows: a named list keyed by reference row id, each list(refMode, members = list(list(row, mode, distance, admittedBy, window, coOccurs))). (conversations/unitConvs grouping is layered on by the caller/server, mirroring tma.conversations().)

Details

Unlike tma.conversations() (which predates the window/weight matrix and uses a single scalar window), this derives windows from the resolved tensor persisted on the accumulated object (x$`_function.params`$tensor) and calls the libqe primitive apply_tensor_members() — the SAME code path the network uses — so the Data View can't drift from the network.