Extract fixed-width result tables from one Mplus section's lines
Source:R/mplus.R
dot-mplus_output_tables.RdHandles Mplus's two real table shapes, found across the local corpus: (1)
a "Model Results"-style table with a stat-vocabulary column header
(.mplus_read_grouped_table()) over rows split into named groups (e.g.
"Within Level" / "Between Level", then within each " IW |" / "
Variances" / " Residual Variances"); and (2) a correlation/covariance
matrix table with a variable-name header and an underline row
(.mplus_read_matrix_table()). The two shapes are genuinely
distinguishable only by trying the stat-vocabulary header match first,
since a bare group-label line (" SW WITH") and a real one-line
header are otherwise indistinguishable by shape alone – see
.mplus_is_stat_header_line().
Arguments
- lines
character vector: one section's lines (from
.mplus_sections())
Value
a list of list(title, data), one per detected table, plus a
consumed logical vector (same length as lines) marking every line
used by a successfully-extracted table – passed to
.mplus_output_labelvalue() so it never re-parses (and mangles) rows
already captured here, e.g. a "MODEL RESULTS" data row like "IW
0.151 0.022 6.887 0.000" otherwise also matches the generic
label/one-trailing-value regex on its LAST column alone.