Roll trajectory emergence up to the authors who own it

The actor-level companion to sniff_trajectory_dynamics(), after the player scoring of Garner et al. (2017): it credits each author with the emergence of the living trajectories their documents belong to. An author who publishes many documents inside high-emergence_index centrals is an owner of the field's emergence; one whose work sits in mature or absorbed lineages is not.

Usage

sniff_trajectory_emergence_owners(
  flow,
  dynamics,
  authors,
  min_docs = 1,
  by_trajectory = FALSE,
  top_n = NULL
)

Arguments

flow: A sniff_trajectory_braid() object.
dynamics: A sniff_trajectory_dynamics() tibble (supplies traj_id and emergence_index).
authors: A long data frame with columns document_id and author, one row per document-author pair.
min_docs: Drop authors with fewer than this many credited documents: summed across trajectories when aggregated, or within the trajectory when by_trajectory = TRUE. Default 1 keeps all. Use a higher cut to read norm/share without one-paper noise.
by_trajectory: If TRUE, return the per-trajectory breakdown (one row per trajectory-author) instead of the field-level aggregate (one row per author). Default FALSE.
top_n: Keep only the highest-scoring authors – overall (by total) when aggregated, or per trajectory (by contribution) when by_trajectory = TRUE. Default NULL keeps all.

Value

A tibble.

When by_trajectory = FALSE (default): one row per author, sorted by descending total – author, total (emergence ownership), ndocs (credited document-trajectory incidences), norm (total / sqrt(ndocs)).

When by_trajectory = TRUE: one row per trajectory-author, sorted by trajectory then descending contribution – traj_id, group, author, ndocs (the author's distinct documents in the trajectory, $n_{a,t}$), emergence_index (the trajectory's weight), contribution (emergence_index * ndocs, the author's share of that trajectory's emergence), and share (ndocs over the trajectory's distinct documents – the author's coverage of its papers).

Details

For every central trajectory with a defined emergence_index (the living population), each author is credited emergence_index(t) per distinct document they hold in t. Absorbed lineages carry no emergence_index and are skipped, so a contributor to a dead tributary is never scored as owning emergence.

$$Total(a) = \sum_t emergence\_index(t)\, n_{a,t}, \qquad Norm(a) = Total(a) / \sqrt{\sum_t n_{a,t}}$$

where $n_{a,t}$ is author $a$'s distinct documents in trajectory $t$. Total favours prolific owners; Norm (per square-root of output, Garner's normalization) surfaces high-intensity authors with fewer papers.

With by_trajectory = TRUE the per-trajectory term $n_{a,t}$ is returned instead of being summed away: one row per (trajectory, author), so you can read which authors dominate each living trajectory rather than the field as a whole. top_n then trims to the leading authors per trajectory.

Usage

Arguments

Value

Details

See also