Package 'blockr.dplyr'

Title: Interactive 'dplyr' Data Transformation Blocks
Description: Extends 'blockr.core' with interactive blocks for visual data wrangling using 'dplyr' and 'tidyr' operations. Users can build data transformation pipelines through a graphical interface without writing code directly. Includes blocks for filtering, selecting, mutating, summarizing, joining, and arranging data, with support for complex expressions, grouping operations, and real-time validation.
Authors: Christoph Sax [aut, cre] (ORCID: <https://orcid.org/0000-0002-7192-7044>), Nicolas Bennett [aut], David Granjon [aut], Mike Page [aut], Bristol Myers Squibb [fnd]
Maintainer: Christoph Sax <[email protected]>
License: GPL (>= 3)
Version: 0.1.0.9001
Built: 2026-05-26 15:07:35 UTC
Source: https://github.com/bristolmyerssquibb/blockr.dplyr

Help Index

HTML dependency for blockr-blocks.css (shared block layout styles)

Description

Exported for reuse by other blockr packages that reuse the shared block-container / row / popover styles.

Usage

blockr_blocks_css_dep()

Value

An htmltools::htmlDependency.

HTML dependency for blockr-core.js (namespace + shared utilities)

Description

Exported for reuse by other blockr packages that build on the shared JS namespace (e.g. blockr.dm).

Usage

blockr_core_js_dep()

Value

An htmltools::htmlDependency.

HTML dependency for blockr-input component (code autocomplete)

Description

Exported for reuse by other blockr packages that embed the shared expression-input component.

Usage

blockr_input_dep()

Value

An htmltools::tagList of htmlDependency objects.

HTML dependency for blockr-select component

Description

Exported for reuse by other blockr packages that embed the shared select component (e.g. blockr.dm's table pickers).

Usage

blockr_select_dep()

Value

An htmltools::tagList of htmlDependency objects.

Build a lightweight column summary for pickers (name + label)

Description

Returns one entry per column: list(name, label). Skips the type detection and anyNA() sweep that build_column_summary() does — those are only needed by the filter block's type-aware condition UI. All other picker blocks only need the name for the option value and the optional label for the secondary text slot, so this helper is the lean path.

Usage

build_column_picker_meta(df)

Arguments

df

A data frame

Details

label is read from attr(col, "label") (e.g. ADaM datasets). It is an empty string when the attribute is missing or not a length-1 character.

Value

A list of list(name, label) entries, one per column.

Build lightweight column summary for JS (names + types only)

Description

Returns name, type, hasNA, and label for each column — no unique values. Sent on data arrival so the filter UI can render column dropdowns instantly. Unique values are fetched on-demand via build_column_values().

Usage

build_column_summary(df)

Arguments

df

A data frame

Details

label is read from attr(col, "label") (e.g. ADaM datasets). It is an empty string when the attribute is missing or not a length-1 character.

Exported as part of the public API so other blockr packages (e.g. blockr.dm) can reuse the same column-metadata protocol.

Value

A list of column summary objects

Build full column metadata for a single column

Description

Computes type, range (numeric), unique values, NA/empty presence, and label for one column. Called on-demand when the user selects a column in the filter block.

Usage

build_column_values(df, col)

Arguments

df

A data frame
col

Column name (character)

Details

Exported as part of the public API so other blockr packages (e.g. blockr.dm) can reuse the same column-metadata protocol.

Value

A column info object (list)

Column Picker UI / Server

Description

Simple selectize-based column picker module used by downstream packages (e.g. blockr.extra).

Usage

column_picker_ui(id, label = "Columns", choices = NULL, selected = NULL)

column_picker_server(id, get_choices, initial_value = NULL)

Arguments

id

Module namespace id
label

Label for the input
choices

Initial choices
selected

Initial selection
get_choices

Reactive returning the available column names.
initial_value

ReactiveVal holding the initial selection.

Value

column_picker_ui returns a tagList; column_picker_server returns a reactive holding the current selection.

HTML dependency for filter block JS + CSS

Description

Exported for reuse by blockr packages that embed the filter block's JS UI (e.g. blockr.dm's dm filter block).

Usage

filter_block_dep()

Value

An htmltools::tagList of htmlDependency objects.

Build a dplyr::filter expression from JS conditions

Description

Exported as part of the public API so other blockr packages (e.g. blockr.dm) can reuse the same filter-condition → expression protocol.

Usage

make_filter_expr(conditions, operator = "&", preserve_order = FALSE)

Arguments

conditions

List of condition objects from JS. Each has a type ("values", "numeric", or "expr") and type-specific fields.
operator

Global operator: "&" or "|"
preserve_order

Reserved; currently unused at this layer.

Value

A language object like dplyr::filter(.(data), ...)

Arrange block (JS-driven)

Description

Sort/arrange block with dynamic rows. Each row has a column picker and a direction toggle (ascending/descending). Auto-submits on any change.

Usage

new_arrange_block(state = list(columns = list()), ...)

Arguments

state

List with columns (list of objects with column and direction fields, where direction is "asc" or "desc").
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_arrange_block(
      state = list(
        columns = list(
          list(column = "Sepal.Length", direction = "desc")
        )
      )
    ),
    data = list(data = iris)
  )
}

Bind cols block (JS-driven, variadic)

Description

Combine multiple data frames side-by-side using dplyr::bind_cols. Minimal UI with no user-configurable parameters. Variadic block: accepts any number of data inputs via the ...args pattern.

Usage

new_bind_cols_block(...)

Arguments

...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_bind_cols_block(),
    data = list(
      a = data.frame(x = 1:3),
      b = data.frame(y = 4:6)
    )
  )
}

Bind rows block (JS-driven, variadic)

Description

Stack multiple data frames on top of each other using dplyr::bind_rows. Simple UI with an optional text input for .id column name. Variadic block: accepts any number of data inputs via the ...args pattern.

Usage

new_bind_rows_block(state = list(id_name = ""), ...)

Arguments

state

List with id_name (string, optional .id column name).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_bind_rows_block(
      state = list(id_name = "source")
    ),
    data = list(a = iris[1:5, ], b = iris[6:10, ])
  )
}

Filter block (JS-driven)

Description

Unified filter block combining simple mode (column + operator + values) and expression mode (free R code) in a single JS-driven UI.

Usage

new_filter_block(state = list(conditions = list(), operator = "&"), ...)

Arguments

state

List with conditions (array of condition objects) and operator ("&" or "|"). Condition types:

  • values: column, values (character vector), mode ("include"/"exclude")

  • numeric: column, op (">", ">=", "<", "<=", "is", "is not"), value

  • expr: expr (R expression as string)

...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_filter_block(
      state = list(
        conditions = list(
          list(type = "values", column = "Species",
               values = list("setosa"), mode = "include")
        ),
        operator = "&"
      )
    ),
    data = list(data = iris)
  )
}

Join block (JS-driven)

Description

Binary join block combining key-based joins (equi and non-equi) with expression mode (free R code for join conditions) in a single JS-driven UI. Takes two data inputs (x and y).

Usage

new_join_block(
  state = list(type = "left_join", keys = list(), exprs = list(), suffix_x = ".x",
    suffix_y = ".y"),
  ...
)

Arguments

state

List with:

  • type: join function name ("left_join", "inner_join", "right_join", "full_join", "semi_join", "anti_join")

  • keys: list of key objects, each with xCol, op, yCol

  • exprs: character vector of R expression strings for join_by()

  • suffix_x: suffix for x columns (default ".x")

  • suffix_y: suffix for y columns (default ".y")

...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_join_block(
      state = list(
        type = "left_join",
        keys = list(
          list(xCol = "id", op = "==", yCol = "id")
        ),
        exprs = list(),
        suffix_x = ".x",
        suffix_y = ".y"
      )
    ),
    data = list(x = iris, y = iris)
  )
}

Mutate block (JS-driven)

Description

JS-driven mutate block with dynamic mutation rows of name + expression pairs. Each mutation defines a new or modified column using an R expression.

Usage

new_mutate_block(
  state = list(mutations = list(list(name = "", expr = "")), by = list()),
  ...
)

Arguments

state

List with mutations (array of objects with name and expr strings) and optional by (array of grouping column names). Each mutation becomes a dplyr::mutate() argument.
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_mutate_block(
      state = list(
        mutations = list(
          list(name = "Sepal.Ratio", expr = "Sepal.Length / Sepal.Width")
        )
      )
    ),
    data = list(data = iris)
  )
}

Pivot longer block (JS-driven)

Description

Reshape data from wide to long format using tidyr::pivot_longer. Multi-select column picker with text inputs for names_to, values_to, names_prefix, and a toggle for values_drop_na. Auto-submits on any change.

Usage

new_pivot_longer_block(
  state = list(cols = list(), names_to = "name", values_to = "value", values_drop_na =
    FALSE, names_prefix = ""),
  ...
)

Arguments

state

List with cols (character vector of columns to pivot), names_to (string), values_to (string), values_drop_na (logical), and names_prefix (string).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_pivot_longer_block(
      state = list(
        cols = list("Sepal.Length", "Sepal.Width"),
        names_to = "measurement",
        values_to = "value",
        values_drop_na = FALSE,
        names_prefix = ""
      )
    ),
    data = list(data = iris)
  )
}

Pivot wider block (JS-driven)

Description

Reshape data from long to wide format using tidyr::pivot_wider. Multi-selects for names_from, values_from, and id_cols (optional). Text inputs for values_fill, names_sep, names_prefix. Auto-submits on any change.

Usage

new_pivot_wider_block(
  state = list(names_from = list(), values_from = list(), id_cols = list(), values_fill =
    NULL, names_sep = "_", names_prefix = ""),
  ...
)

Arguments

state

List with names_from (character vector), values_from (character vector), id_cols (character vector, optional), values_fill (value or NULL), names_sep (string), and names_prefix (string).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_pivot_wider_block(
      state = list(
        names_from = list("Species"),
        values_from = list("Sepal.Length"),
        id_cols = list(),
        values_fill = NULL,
        names_sep = "_",
        names_prefix = ""
      )
    ),
    data = list(data = iris)
  )
}

Rename block (JS-driven)

Description

Column rename block with dynamic rows. Each row maps an old column name to a new name. Auto-submits on any change (300ms debounce for text input).

Usage

new_rename_block(state = list(renames = list()), ...)

Arguments

state

List with renames (named list where names are new column names and values are old column names).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_rename_block(
      state = list(
        renames = list(sepal_len = "Sepal.Length")
      )
    ),
    data = list(data = iris)
  )
}

Select block (JS-driven)

Description

Column selection block with reorderable multi-select, exclude toggle, and distinct toggle. Auto-submits on any change.

Usage

new_select_block(
  state = list(columns = list(), exclude = FALSE, distinct = FALSE),
  ...
)

Arguments

state

List with columns (character vector of column names), exclude (logical), and distinct (logical).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_select_block(
      state = list(
        columns = list("Sepal.Length", "Species"),
        exclude = FALSE,
        distinct = FALSE
      )
    ),
    data = list(data = iris)
  )
}

Separate block (JS-driven)

Description

Split a single column into multiple columns using tidyr::separate. Single select for source column, text input for into (comma-separated new column names), text input for sep, toggle pills for remove and convert. Auto-submits on any change.

Usage

new_separate_block(
  state = list(col = "", into = list(), sep = "[^[:alnum:]]+", remove = TRUE, convert =
    FALSE, extra = "warn", fill = "warn"),
  ...
)

Arguments

state

List with col (string, source column), into (character vector of new column names), sep (string or regex), remove (logical), convert (logical), extra (string), and fill (string).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  df <- data.frame(x = c("a-1", "b-2", "c-3"), stringsAsFactors = FALSE)
  serve(
    new_separate_block(
      state = list(
        col = "x",
        into = list("letter", "number"),
        sep = "-",
        remove = TRUE,
        convert = FALSE,
        extra = "warn",
        fill = "warn"
      )
    ),
    data = list(data = df)
  )
}

Slice block (JS-driven)

Description

Subset rows by position using dplyr::slice_* family of functions. Single select for type (head/tail/min/max/sample), number input for n, optional column selects for order_by and weight_by, toggle pills for with_ties and replace, multi-select for grouping (.by). Auto-submits on any change.

Usage

new_slice_block(
  state = list(type = "head", n = 5L, prop = NULL, order_by = "", with_ties = TRUE,
    weight_by = "", replace = FALSE, rows = "1:5", by = list()),
  ...
)

Arguments

state

List with type (string: "head", "tail", "min", "max", "sample"), n (integer), prop (numeric or NULL), order_by (string), with_ties (logical), weight_by (string), replace (logical), and by (character vector).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_slice_block(
      state = list(
        type = "head",
        n = 10L,
        prop = NULL,
        order_by = "",
        with_ties = TRUE,
        weight_by = "",
        replace = FALSE,
        by = list()
      )
    ),
    data = list(data = iris)
  )
}

Summarize block (JS-driven)

Description

JS-driven summarize block with two row types: simple (function + column) and expression (free R code). Includes a group-by section for grouped summaries.

Usage

new_summarize_block(state = list(summaries = list(), by = list()), ...)

Arguments

state

List with summaries (array of summary objects) and by (character vector of grouping columns). Summary types:

  • simple: name, func (e.g. "mean"), col (column name)

  • expr: name, expr (R expression string)

...

Additional arguments forwarded to blockr.core::new_block()

Extending available functions

The list of available summary functions can be extended using the blockr.dplyr.summary_functions option. Set this option to a named character vector where names are display labels and values are function calls with proper namespacing: 

options(
  blockr.dplyr.summary_functions = c(
    "extract parentheses (paren_num)" = "blockr.topline::paren_num"
  )
)

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_summarize_block(
      state = list(
        summaries = list(
          list(type = "simple", name = "avg_sl", func = "mean",
               col = "Sepal.Length")
        ),
        by = list("Species")
      )
    ),
    data = list(data = iris)
  )
}

Unite block (JS-driven)

Description

Paste together multiple columns into one using tidyr::unite. Text input for new column name, multi-select for columns to unite, text input for separator, toggle pills for remove and na.rm. Auto-submits on any change.

Usage

new_unite_block(
  state = list(col = "united", cols = list(), sep = "_", remove = TRUE, na_rm = FALSE),
  ...
)

Arguments

state

List with col (string, new column name), cols (character vector of columns to unite), sep (string), remove (logical), and na.rm (logical).
...

Additional arguments forwarded to blockr.core::new_block()

Examples

if (interactive()) {
  library(blockr.core)
  serve(
    new_unite_block(
      state = list(
        col = "full_name",
        cols = list("Sepal.Length", "Sepal.Width"),
        sep = "_",
        remove = TRUE,
        na.rm = FALSE
      )
    ),
    data = list(data = iris)
  )
}