Section author: Jonathon Love
Assign results to tables (in R
)
This document describes the properties and methods of a jamovi table object.
The values of properties can be accessed using the $
operator,
followed by the name. For example, to retrieve the title of a table, one
can go:
table$title
The methods of a table object are called using the $
operator as
well. For example:
table$setRow(rowKey=1, values=list(t=2.3, df=2, p=0.45))
Properties
name
a string specifying the name of the table
title
a string specifying the title of the table
visible
TRUE
or FALSE
, whether the table is visible or not
status
a string, one of 'complete'
, 'error'
, 'inited'
, 'running'
rowKeys
a list of ‘keys’
state
the state
Methods
setStatus(status)
sets the table’s status, should be one of 'complete'
, 'error'
,
'inited'
, 'running'
setVisible(visible=TRUE)
overrides the tables default visibility
setTitle(title)
sets the table’s title
setError(message)
sets the table’s status to ‘error’, and assigns the error message
setState(object)
sets the state object on the table
addColumn(name, …)
adds a new column to the table, the following arguments are possible:
argument |
type |
details |
---|---|---|
|
string |
the name of the column |
|
integer |
the index to insert the column at. if unspecified, the column is appended. |
|
string |
the title to appear at the top of the column. if unspecified, the name is used. |
|
string |
the title to appear above column titles |
|
|
whether the column should be visible. if a string is specified, this must be a data-binding to an option. |
|
string |
either a string that will be placed in every cell, or a data-binding |
|
string |
‘integer’, ‘number’ or ‘text’; text is left aligned, numbers are right aligned, integers are formatted to zero decimal places |
|
string |
a comma separated list of values, such as ‘zto’, ‘pvalue’ |
|
|
if TRUE, when cells in the column are contiguous, and contain the same value, the lower cells will be made blank. |
addRow(rowKey, values)
Adds a row to the table. rowKey
is an object which uniquely
identifies the row – for many cases, simply providing the index is
sufficient. values
is a named list with the values to place in the
cells of that row. The names must correspond to the column names. Not
all column values must be provided, and if a blank row is desired, the
values argument can be omitted entirely.
deleteRows()
Deletes all the rows in the table
setRow(rowKey, values)
Sets the values in an existing row. rowKey
is a key uniquely
identifying the row, and values
is a named list of values. The names
must correspond to the column names. Not all column values need to be
provided.
Warning
Note that you must explicitly name the rowKey argument:
setRow(rowKey=...)
to differentiate from the rowNo=...
argument below
setRow(rowNo, values)
Sets the values in an existing row. rowNo
is a number specifying the
row number, and values
is a named list of values. The names must
correspond to the column names. Not all column values need to be
provided.
Warning
Note that you must explicitly name the rowNo argument when calling this method:
setRow(rowNo=...)
to differentiate from the rowKey=...
argument above
addFormat(rowKey, col, format)
addFormat(rowNo, col, format)
Adds additional formatting to a cell. col
can be an index or a name.
format can be one of:
Cell.BEGIN_GROUP
Cell.END_GROUP
Cell.BEGIN_END_GROUP
Cell.NEGATIVE
Cell.BEGIN_GROUP adds additional padding above the cell. Cell.END_GROUP adds additional padding below the cell. Cell.NEGATIVE colours the value red.
setCell(rowKey, col, value)
setCell(rowNo, col, value)
Sets the value of a cell. Generally setRow() is more efficient.
getCell(rowKey, col)
getCell(rowNo, col)
Retrieves a cell.
addFootnote(rowKey, col, note)
addFootnote(rowNo, col, note)
Adds a footnote to the cell.
addSymbol(rowKey, col, symbol)
addSymbol(rowNo, col, symbol)
Adds a symbol to a cell – for example an asterisk denoting significance.
setNote(key, note, init=TRUE)
setNote() adds (or clears) a note placed in the footer of the table.
key
: a string identifying the notenote
: a string representing the text of the note (or NULL)init
: whether this be considered aninit
note
Specifying a note
of NULL causes the note to be removed.
init
notes are those that are added during the init phase.
init
notes are typically based on the values of the options. For
example, if the user has specified an alternative hypothesis — that
population one is greater than population two — the analysis could add a
note indicating this. In contrast, non-init
notes are created in the
run phase. An example might be the number of subjects that were
excluded from the analysis as a result of containing missing values.
init
notes are typically based on the values of options, where as
non-init
notes depend on the data.
In practice, when an analysis is changed or re-run, init
notes are
not restored from state; they are simply recreated during the init
phase. In contrast, non-init
notes are restored from state.
Note that if the text of the note will always be the same, it is
recommended to set the note in the .r.yaml
file instead.