Package 'spreadview'

Compose a Spreadsheet with Multiple Frequency Tables

Description

Creates an Excel workbook containing frequency tables for multiple variables, optionally grouped and weighted. Each variable's table is row-bound together into a single sheet. When grouping is specified, cells are colored based on adjusted standardized residuals.

Usage

compose_spreadsheet(
  data,
  vars,
  group = NULL,
  weight = NULL,
  file = NULL,
  na.rm = TRUE,
  percent = TRUE
)

Arguments

data	A data frame or data.table containing the variables to tabulate.
vars	Character vector of variable names to create tables for. Each variable will be passed to compose_table().
group	Character vector specifying grouping variables. Default is NULL for no grouping. Passed to compose_table().
weight	Character string naming the weight variable to use for weighted frequencies. Passed to compose_table().
file	Character string specifying the output file path. If NULL (default), returns the workbook object without saving.
na.rm	Should missing values be removed from the table? Defaults to TRUE.
percent	Logical. If TRUE (default), percentage values are formatted as percentages (e.g., 50\ as plain numbers multiplied by 100 (e.g., 50).

Details

The function iterates over each variable in vars, applies compose_table() with the specified parameters, and combines the results using data.table::rbindlist(). The resulting spreadsheet is added to a new workbook with openxlsx2.

The header spans two rows:

• Row 1: Column names for non-group columns (e.g., "label", "n", "total") and group variable names for grouped columns (e.g., "IDE_8")

• Row 2: Empty strings for non-group columns, category names for grouped columns (e.g., "muž", "žena")

When grouping variables are specified, cells are colored based on adjusted standardized residuals:

• Blue shades for positive residuals (more than expected)

• Red shades for negative residuals (less than expected)

• Color intensity increases with absolute residual value

• Residuals beyond ±2 indicate significance at ~0.05 level

Value

An openxlsx2 workbook object containing a single worksheet with all frequency tables row-bound together. Cells are colored based on adjusted standardized residuals when grouping is used.

See Also

compose_table, compute_adj_residuals

Examples

## Not run: 
compose_spreadsheet(data = penguins, var = c("species", "sex"), group = "island")

## End(Not run)

---

Compose a frequency table with optional grouping

Description

Creates a frequency table for a categorical variable, optionally cross-tabulated by one or more grouping variables. The function automatically adds variable labels and standardizes column names for easier downstream processing.

Usage

compose_table(
  data,
  var,
  group = NULL,
  weight = NULL,
  prop = "none",
  na.rm = TRUE
)

Arguments

data	A data frame or data.table containing the variables to tabulate.
var	Character string specifying the name of the variable to tabulate.
group	Character vector of variable name(s) to group by. If NULL (default), creates a simple frequency table. If provided, creates cross-tabulations with each grouping variable. Group columns are prefixed with ⁠[group_name]⁠ in the output.
weight	Character string specifying the name of the weight variable to use for weighted frequencies.
prop	Character string specifying proportion type. Options are "none" for counts only, "col" for column proportions, or "row" for row proportions. Default is "none".
na.rm	Should missing values be removed from the table? Defaults to TRUE.

Value

A data.table with the following structure:

item	Variable name of the tabulated variable
label	Variable label from the var attribute (empty string if not available)
var	Factor levels of the tabulated variable
n	Weighted counts
total	Overall proportions (if prop != "none")
\link{group} level	One column per level of each grouping variable (if group is specified)

Examples

compose_table(penguins, var = "species", group = "island", prop = "col")

---

Compute adjusted standardized residuals for contingency tables

Description

Calculates adjusted standardized residuals for a contingency table formed by crossing a row variable with one or more grouping (column) variables. These residuals measure how much each observed cell frequency deviates from what would be expected under independence, adjusted for both row and column proportions.

Usage

compute_adj_residuals(data, var, group, weight)

Arguments

data	A data frame or tibble containing the survey data.
var	Character string specifying the row variable name.
group	Character vector specifying one or more column grouping variable names. When multiple groups are provided, residuals are computed separately for each group's contingency table.
weight	Character string specifying the weight variable name.

Details

Adjusted standardized residuals are calculated as:

$\frac{O - E}{\sqrt{E \times (1 - p_{row}) \times (1 - p_{col})}}$

where $O$ is the observed frequency, $E$ is the expected frequency under independence, $p_{row}$ is the row proportion, and $p_{col}$ is the column proportion.

These residuals can be interpreted like z-scores: values beyond ±2 suggest significant deviation from independence at approximately the 0.05 level. They are equivalent to the standardized residuals from chisq.test().

When multiple grouping variables are provided, residuals are computed independently for each two-way contingency table (var × group_i), then combined into a single output table matching the structure of compose_table().

Value

A data.table with the same structure as the output from compose_table() with the same grouping variables. The first column contains the row variable levels, and subsequent columns contain adjusted standardized residuals for each group level. When multiple groups are specified, columns are prefixed with [group_name] to match compose_table() output.

Examples

## Not run: 
# Single group
compute_adj_residuals(cvvm, var = "t_VZD", group = "IDE_8", weight = "soc_weight")

# Multiple groups
compute_adj_residuals(
  cvvm,
  var = "t_VZD",
  group = c("t_VEK_6", "IDE_8"),
  weight = "soc_weight"
)

## End(Not run)

---

Calculate Frequency Tables with Optional Proportions

Description

Computes weighted frequency tables for a variable, optionally grouped by another variable. Can return counts or proportions calculated across different margins (total, row, or column).

Usage

count_freqs(
  data,
  var,
  group = NULL,
  weight = NULL,
  prop = "none",
  na.rm = TRUE,
  include_empty = TRUE
)

Arguments

data	A data frame or object that can be coerced to a data.table.
var	Character string specifying the name of the variable to tabulate.
group	Character string specifying the name of the grouping variable, or NULL (default) for no grouping.
weight	Character string specifying the name of the weight variable to use for weighted frequencies. If no weights are specified, equal weights are assumed.
prop	Character string indicating the type of proportion to calculate. Options are: "none" (default): Return weighted counts only "total": Proportions of the grand total "row": Proportions within each row (within each level of var) "col": Proportions within each column (within each level of group)
na.rm	Logical. If TRUE, removes rows with missing values in var before computing frequencies. Default is FALSE.
include_empty	Logical. If TRUE (default), factor levels of var (and group, if specified) with no observed values are included in the result with n = 0.

Value

A data.table containing:

• When group is NULL: A table with var and either n (counts) or total (proportions)

• When group is specified and prop != "none": A wide-format table with var as rows and group levels as columns, containing proportions

• When group is specified and prop == "none": A long-format table with var, group, and n columns

Examples

count_freqs(data = penguins, var = "species", prop = "none")
count_freqs(data = penguins, var = "species", group = "island", prop = "col")

---

Get names of categorical variables from a data frame

Description

Extracts the names of factor variables from a data frame. Optionally includes character variables and allows exclusion of specific variable names.

Usage

get_categorical_vars(data, exclude = NULL, include_char = FALSE)

Arguments

data	A data frame to extract categorical variable names from.
exclude	A character vector of variable names to exclude from the results. Default is NULL (no exclusions).
include_char	Logical. If TRUE, character variables are included in addition to factor variables. Default is FALSE.

Value

A character vector of variable names.

Examples

# Get only factor variables
get_categorical_vars(iris)

# Include character variables
get_categorical_vars(mtcars, include_char = TRUE)

# Exclude specific variables
get_categorical_vars(iris, exclude = c("Species"))

---