Tables

table.aggregate#

Rows grouped by the values of some columns and summarised, as dplyr's group_by and summarise: a row per group (the groups in the order of their values), its columns n (the rows in the group) and, for every column summarised and statistic, column_statistic (area_sum, area_mean). Missing values are left out of the statistics; sd is the sample's (n - 1, as R's sd). Without columns to group by, the whole table is one group.

table.calc#

A column computed row by row from an expression over the others, as a spreadsheet's field calculator or dplyr's mutate: the expressions of map algebra (area / 10000, cells if class == 1 else 0), the node's other inputs as numbers by their names. A column of the same name is replaced, else it is added at the end.

Takes any other inputs by name (any).

table.filter#

The rows of a table where an expression is true, as dplyr's filter: the columns by their names, the expressions of map algebra (class == 3 and area > 10; isnull(x) for a missing value), the node's other inputs as numbers by their names. A row where the expression is missing is left out.

Takes any other inputs by name (any).

table.join#

Two tables joined by key columns, as dplyr's joins: every row of the left with each row of the right of the same key (left: the left's rows without one kept, the right's columns missing; inner: only the rows of keys in both; full: the right's rows without one added too). The right's other columns follow the left's; one of a name the left has takes a suffix.

table.of#

A table written in the settings: its columns' names and its rows of numbers (null: no value), such as the small tables a planning problem takes.

table.pivot#

A table from long to wide or back, as tidyr's pivot_wider and pivot_longer. Wider: a row per value of the id columns, a column per value of the names column (named prefix and the value: to_1, to_2...) holding the values column (transitions from, to, cells into a matrix). Longer: the columns given (or every column but the id ones) into rows of two columns, the names and the values; a table's values being numbers, the name of a column is the number in it (2011 of y2011 with prefix y; 1 of to_1), else its place among them (1, 2...).

table.read#

A table from a file: comma-separated values (.csv), tab-separated (.tsv, .txt), with ; as Excel writes them where a comma marks the decimals (the separator and the decimal mark found by themselves), or JSON (.json, a list of rows). A header names the columns; an empty value, or NA, is none. The columns of text (names, not numbers) are left out: a table is of numbers.

table.select#

Columns of a table chosen (in the order given), left out, or renamed, as dplyr's select and rename.

table.sort#

The rows of a table sorted by some columns (the first, then the next where it ties; missing values last), as dplyr's arrange, and the first rows kept (top: the ten largest patches, say).

table.write#

Write a table to a file, its format by the extension: comma-separated values (.csv, the default), tab-separated (.tsv, .txt) or JSON (.json, a list of rows); a value missing is left empty (null in JSON). For a spreadsheet that reads a comma as the decimal mark: separator ; and decimal ,.