Package 'DataSum'

DataSum: Modern Data Summaries and Diagnostic Reports

Description

DataSum provides robust, NA-aware summaries, data profiles, normality diagnostics, outlier checks, reproducible report scaffolding, and an optional Shiny app for statistical data inspection.

See Also

summarize_vector(), summarize_data(), profile_data(), datasum_report(), run_datasum_app()

---

Create a DataSum Diagnostic Report

Description

Creates a self-contained Quarto source file containing dataset diagnostics, variable summaries, warnings, and formula definitions. Set render = TRUE to render the report through the optional quarto package.

Usage

datasum_report(
  data,
  path = NULL,
  format = c("qmd", "html", "pdf", "docx"),
  title = "DataSum Diagnostic Report",
  by = NULL,
  alpha = 0.05,
  digits = 3,
  render = FALSE
)

Arguments

data	A data frame or tibble.
path	Output path. When render = FALSE, this should usually end in .qmd. If omitted, a temporary .qmd file is created.
format	One of "qmd", "html", "pdf", or "docx".
title	Report title.
by	Optional grouping columns passed to profile_data().
alpha	Significance level for normality decisions.
digits	Optional number of digits used to round numeric output.
render	Logical; if TRUE, render the Quarto document. Rendering requires the optional quarto package and a working Quarto installation.

Value

The generated file path, invisibly.

Examples

report <- datasum_report(iris, render = FALSE)
file.exists(report)

---

Profile a Data Frame

Description

Builds a dataset-level profile containing variable summaries, dataset shape, missingness, duplicate-row counts, type counts, and warnings that deserve analyst attention.

Usage

profile_data(data, by = NULL, alpha = 0.05, digits = NULL)

Arguments

data	A data frame or tibble.
by	Optional grouping columns passed to summarize_data().
alpha	Significance level for normality decisions.
digits	Optional number of digits used to round numeric output.

Value

A datasum_profile list with dataset, summary, and warnings.

Examples

profile <- profile_data(iris)
profile$dataset

---

Launch the DataSum Shiny App

Description

Opens an interactive Shiny interface for uploading a CSV file, inspecting dataset diagnostics, visualizing a selected variable, and downloading a reproducible DataSum report source file.

Usage

run_datasum_app(data = NULL)

Arguments

data	Optional data frame used as the starting dataset. If omitted, the app starts with datasets::iris until a CSV is uploaded.

Value

A Shiny application object.

Examples

if (requireNamespace("shiny", quietly = TRUE)) {
  app <- run_datasum_app(iris)
}

---

Summarize a Data Frame

Description

Applies summarize_vector() to every column in a data frame. Optional grouped summaries are supported by passing one or more grouping column names to by.

Usage

summarize_data(data, by = NULL, alpha = 0.05, digits = NULL)

Arguments

data	A data frame or tibble.
by	Optional character vector of grouping columns.
alpha	Significance level for normality decisions.
digits	Optional number of digits used to round numeric output.

Value

A data.frame, one row per summarized variable and group.

Examples

summarize_data(iris)
summarize_data(iris, by = "Species")

---

Summarize a Single Vector

Description

Computes a one-row, NA-aware diagnostic summary for one vector. Numeric vectors receive robust and classical statistics, outlier counts, and a normality diagnostic. Non-numeric vectors receive safe type, missingness, uniqueness, and mode summaries.

Usage

summarize_vector(x, name = NA_character_, alpha = 0.05, digits = NULL)

Arguments

x	A vector.
name	Optional variable name to store in the variable column.
alpha	Significance level for the normality decision. Defaults to 0.05.
digits	Optional number of digits used to round numeric output. By default, numeric values are not rounded.

Value

A one-row data.frame with summary statistics and diagnostics.

Examples

summarize_vector(c(1, 2, 2, NA, 5), name = "score")
summarize_vector(factor(c("control", "treatment", "control")))

---