Package 'ncodeR'

Title: Techniques for Automated Classifiers
Description: A set of techniques that can be used to develop, validate, and implement automated classifiers. A powerful tool for transforming raw data into meaningful information, 'ncodeR' (Shaffer, D. W. (2017) Quantitative Ethnography. ISBN: 0578191687) is designed specifically for working with big data: large document collections, logfiles, and other text data.
Authors: Cody L Marquart [aut, cre] , Zachari Swiecki [aut], Brendan Eagan [aut], David Williamson Shaffer [aut]
Maintainer: Cody L Marquart <[email protected]>
License: GPL-3 | file LICENSE
Version: 0.2.1.1
Built: 2026-09-15 09:17:08 UTC
Source: https://gitlab.com/epistemic-analytics/qe-packages/ncoder

Help Index

Code to data.frame

Description

Code to data.frame

Usage

## S3 method for class 'Code'
as.data.frame(x, row.names = NULL, optional = FALSE, ...)

Arguments

x

Code object to convert
row.names

NULL or a character vector giving the row names for the data frame. Missing values are not allowed.
optional

logical. If TRUE, setting row names and converting column names
...

additional arguments to be passed to or from methods

Value

data.frame

Examples

data(RS.data)
rs = RS.data
newcode = create.code(name = "Data", expressions = c("number","data"), excerpts = rs$text)
as.data.frame(newcode)

CodeSet to data.frame

Description

CodeSet to data.frame

Usage

## S3 method for class 'CodeSet'
as.data.frame(x, row.names = NULL, optional = FALSE, ...)

Arguments

x

CodeSet to convert
row.names

NULL or a character vector giving the row names for the data frame. Missing values are not allowed.
optional

logical. If TRUE, setting row names and converting column names
...

additional arguments to be passed to or from methods

Value

data.frame

Examples

data(RS.data)
rs = RS.data
newcode = create.code(name = "Data", expressions = c("number","data"), 
    excerpts = rs$text)
code.set = code.set("Demo RS CodeSet", "CodeSet made for the demo", 
    excerpts = rs$text, codes = c(newcode))
as.data.frame(code.set)

Match a list of expressions against some set of excerpts

Description

Autocodes all codes provided, either directly with code or as part of a provided codeset

Usage

autocode(
  x = NULL,
  expressions = NULL,
  excerpts = NULL,
  simplify = T,
  mode = "all"
)

Arguments

x

Object to autocode. Either a Code or CodeSet
expressions

Expressions to use for coding (optional)
excerpts

Excerpts to code
simplify

If TRUE, returns a data.frame, else returns a Code or CodeSet object
mode

Either all, training, or test representing the set of excerpts that should be recoded in the computerSet

Value

data.frame of is simplify = T (default), otherwise the Code or CodeSet object with updated computerSets

ncodeR Code object

Description

ncodeR Code object

ncodeR Code object

Value

Code object Create Code

Public fields

call

TBD

name

TBD

definition

TBD

codeSet

TBD

testSet

TBD

trainingSet

TBD

computerSet

TBD

ignoredSet

TBD

examples

TBD

excerpts

TBD

holdoutExcerpts

TBD

touchableExcerpts

TBD

touchedIndices

TBD

additionalExcerpts

TBD

secondRaterSet

TBD

statistics

TBD

baserate

TBD

baserateInflation

TBD

Methods

Public methods

Method new()

Usage
Code$new(
  name,
  definition,
  testSet = NULL,
  trainingSet = NULL,
  computerSet = NULL,
  secondRaterSet = NULL,
  ignoredSet = c(),
  examples = NULL,
  excerpts = NULL,
  holdoutSize = 0.9,
  ...
)
Arguments
name

TBD

definition

TBD

testSet

TBD

trainingSet

TBD

computerSet

TBD

secondRaterSet

TBD

ignoredSet

TBD

examples

TBD

excerpts

TBD

holdoutSize

TBD

...

TBD

Returns

Code object Function to override

Method process()

Usage
Code$process()
Returns

Default, error, unless overridden Kappa on Code

Method kappa()

Usage
Code$kappa(which = c("training", "test"))
Arguments
which

TBD

Returns

double Differences in sets

Method differences()

Usage
Code$differences(data = NULL, col1 = NULL, col2 = NULL, cols = NULL)
Arguments
data

TBD

col1

TBD

col2

TBD

cols

TBD

Returns

vector of indices Clear TestSet

Method clearTestSet()

Usage
Code$clearTestSet()
Returns

NULL Concatenate Expressions

Method concat()

Usage
Code$concat()
Returns

character of concatenated expressions Get Value

Method getValue()

Usage
Code$getValue(wh)
Arguments
wh

character

Returns

value stored at wh Set Value

Method setValue()

Usage
Code$setValue(wh, val)
Arguments
wh

character

val

object

Method clone()

The objects of this class are cloneable with this method.

Usage
Code$clone(deep = FALSE)
Arguments
deep

Whether to make a deep clone.

Create CodeSet

Description

Create a new CodeSet object

Usage

code.set(title = "", description = "", excerpts = c(), codes = c())

Arguments

title

Title for the CodeSet
description

Description of the CodeSet
excerpts

Set of excerpts to use with the CodeSet
codes

Set of codes to attach to the CodeSet

Value

CodeSet object

Examples

data(RS.data)
rs = RS.data
code.set = code.set("Demo RS CodeSet", "CodeSet made for the demo", excerpts = rs$text, codes = c())

CodeSet

Description

Object representing a set of codes

Value

CodeSet object Create CodeSet

Public fields

title

Title of the CodeSet

call

TBD

codes

list of Codes

description

String description of the set of codes to be included

excerpts

Character vector of text excerpts to code (optional)

Methods

Public methods

Method new()

Usage
CodeSet$new(title = "ne", description = NULL, excerpts = c(), codes = c())
Arguments
title

TBD

description

TBD

excerpts

TBD

codes

TBD

Returns

CodeSet

Method clone()

The objects of this class are cloneable with this method.

Usage
CodeSet$clone(deep = FALSE)
Arguments
deep

Whether to make a deep clone.

Examples

data(RS.data)
rs = RS.data
code.set = code.set("Demo RS CodeSet", "CodeSet made for the demo", excerpts = rs$text, codes = c())

Create a code

Description

Create a code

Usage

create.code(
  name = "NewCode",
  definition = NULL,
  excerpts = NULL,
  type = "Regex",
  ...
)

Arguments

name

Name of the code
definition

Definition of the Code
excerpts

Character vectore of excerpts to use for Coding
type

Character string representing the type of code (Default: "Regex")
...

Additional parameters

Value

Code object

Examples

data(RS.data)
rs = RS.data

# Generate a Code
newcode = create.code(name = "Data", expressions = c("number","data"), excerpts = rs$text)

Find Differences

Description

Find rows that differ within a data.frame or two vectors

Usage

differences(code = NULL, wh = "trainingSet", to = "computerSet")

Arguments

code

Code object to search for differences
wh

Set to use as the base comparison
to

Set to compare wh to

Value

logical vector representing indices that are coded differently

vector of indices representing differences

Expression Matching

Description

Match a set of text excerpts against a set of regular expressions

Usage

expression.match(excerpts, expressions, names = list(NULL, "V1"))

Arguments

excerpts

Character vector to match against
expressions

Character vector of expressions
names

Character vector to use for dimension names

Value

Matrix representing matched expressions

Handset indices

Description

Handset indices

Usage

getHandSetIndices(
  codeToUse,
  handSetLength = 20,
  handSetBaserate = 0.2,
  unseen = F
)

Arguments

codeToUse

[TBD]
handSetLength

[TBD]
handSetBaserate

[TBD]
unseen

[TBD]

Get indices to code

Description

Get indices to code

Usage

getHandSetIndices2(
  code,
  handSetLength = 20,
  handSetBaserate = 0.2,
  unseen = FALSE,
  this.set = NULL
)

Arguments

code

Code object
handSetLength

Number of excerpts to put into the test set
handSetBaserate

Minimum number of positives that should be in the test set
unseen

[TBD]
this.set

[TBD]

Value

Code object with an updated test set and computer set

Handcode Code

Description

Handcode a set of excerpts using a vector of expressions

Usage

handcode(
  code = NULL,
  excerpts = NULL,
  expressions = NULL,
  n = ifelse(is.null(this.set), 10, length(this.set)),
  baserate = 0.2,
  unseen = FALSE,
  this.set = NULL,
  results = NULL
)

Arguments

code

Code object to handcode
excerpts

Excerpts to code (optional)
expressions

Expressions to code with (options)
n

Number of excerpts to handcode
baserate

Value between 0 and 1, inflates the baserate chosen excerpts to code, ensuring the number of positive at least equal to n * baserate
unseen

Logical or number Indicating additional excerpts with unseen words should be added. If TRUE (default), two words added or by 'number'
this.set

[TBD]
results

[TBD]

Value

Code

ncodeR for qualitative coding

Description

ncodeR is used for generating codes and coding datasets

Calculate statistics

Description

Run tests (kappa, rho) on the given Code

Usage

old_test(
  code,
  kappaThreshold = 0.65,
  baserateInflation = 0.2,
  type = c("training", "test")
)

Arguments

code

Code object to test
kappaThreshold

Threshold used for calculating rhoR::rho
baserateInflation

inflation rate to use when sampling handsets
type

vector indicating which stats should be calculated

Value

Code object with updated statistics property

Print a Code summary

Description

Print a Code summary

Usage

## S3 method for class 'summary.Code'
print(x, ...)

Arguments

x

list from summary()
...

Additional parameters

Value

Prints code summary

Examples

data(RS.data)
rs = RS.data
newcode = create.code(name = "Data", 
    expressions = c("number","data"), excerpts = rs$text)
summary(newcode)

Print the summary of a CodeSet

Description

Print the summary of a CodeSet

Usage

## S3 method for class 'summary.CodeSet'
print(x, ...)

Arguments

x

Summary of a CodeSet
...

Additional parameters

Value

prints summary

Examples

data(RS.data)
rs = RS.data

newcode = create.code(name = "Data", 
    expressions = c("number","data"), excerpts = rs$text)
code.set = code.set("Demo RS CodeSet", "CodeSet made for the demo", 
    excerpts = rs$text, codes = c(newcode))
summary(code.set)

Print a TestList summary

Description

Print a TestList summary

Usage

## S3 method for class 'summary.TestList'
print(x, ...)

Arguments

x

list from summary()
...

Additional parameters

Value

prints summary

Examples

data(RS.data)
rs = RS.data
newcode <- create.code("Data", expressions = c("number","data"), excerpts = rs$text)
newcode <- handcode(newcode, this.set = 10:15, results = 0)
newcode = test(code = newcode, kappa_threshold = 0.65)
summary(newcode$statistics)

RegexCode

Description

Creates an object for Regular Expression coding. No need to call this directly, create.code is a nice wrapper around this and any other types of Codes

Value

RegexCode object Create RegexCode

Super class

ncodeR::Code -> RegexCode

Public fields

expressions

Character vector of regular expressions

metadata

TBD

Methods

Public methods

Inherited methods

Method new()

Usage
RegexCode$new(name, definition, ..., excerpts = NULL, expressions = NULL)
Arguments
name

TBD

definition

TBD

...

TBD

excerpts

TBD

expressions

TBD

Returns

Process Code

Method process()

Usage
RegexCode$process(excerpts = self$excerpts)
Arguments
excerpts

Options excerpts to code

Returns

Code results Add word

Method add()

Usage
RegexCode$add(word)
Arguments
word

To add to expressions

Returns

NULL Remove word

Method remove()

Usage
RegexCode$remove(word)
Arguments
word

to remove

Returns

NULL Concatenate expressions

Method concat()

Usage
RegexCode$concat()
Returns

character of concatentated expressions

Method clone()

The objects of this class are cloneable with this method.

Usage
RegexCode$clone(deep = FALSE)
Arguments
deep

Whether to make a deep clone.

Examples

data(RS.data)
rs = RS.data

# Generate a Code
newcode = RegexCode$new(name = "New Code", definition = "Some definition", 
    excerpts = rs$text, expressions = c("number","data"))

Resolve differences

Description

Resolve differing results

Usage

resolve(
  code = NULL,
  trainingSet = NULL,
  computerSet = NULL,
  expressions = NULL,
  excerpts = NULL,
  ignored = NULL
)

Arguments

code

Code to resolve coding differences
trainingSet

Optionally provide a trainingSet, default: code$trainingSet
computerSet

Optionally provide a computerSet, default: code$computerSet
expressions

Optionally provide a set of expressions, default: code$expressions
excerpts

Optionally provide a set of excerpts, default: code$excerpts
ignored

Optionally proivde a set of excerpts to ignore during the resolve cycle loop

Rescushell Chat Data

Description

A dataset containing sample chat data from the Rescushell Virtual Internship

Usage

RS.data

Format

An object of class data.frame with 3824 rows and 20 columns.

Obtain summary of a Code object

Description

Obtain summary of a Code object

Usage

## S3 method for class 'Code'
summary(object, ...)

Arguments

object

Code to summarize
...

Additional parameters

Value

List of Code summary

Examples

data(RS.data)
rs = RS.data
newcode = create.code(name = "Data", 
    expressions = c("number","data"), excerpts = rs$text)
summary(newcode)

Obtain a summary of the CodeSet

Description

Obtain a summary of the CodeSet

Usage

## S3 method for class 'CodeSet'
summary(object, ...)

Arguments

object

CodeSet object
...

Additional parameters

Value

list containing description and Code summaries

Examples

data(RS.data)
rs = RS.data

newcode = create.code(name = "Data", 
    expressions = c("number","data"), excerpts = rs$text)
code.set = code.set("Demo RS CodeSet", "CodeSet made for the demo", 
    excerpts = rs$text, codes = c(newcode))
summary(code.set)

Obtain a summary of a Code's test results

Description

Obtain a summary of a Code's test results

Usage

## S3 method for class 'TestList'
summary(object, ...)

Arguments

object

TestList object of Code
...

Additional parameters

Value

list of Test summary

Examples

data(RS.data)
rs = RS.data
newcode = create.code(name = "Data", 
    expressions = c("number","data"), excerpts = rs$text)
newcode <- handcode(newcode, this.set = 10:15, results = 0)
newcode = test(code = newcode, kappa_threshold = 0.65)
summary(newcode$statistics)

Test code

Description

Test code

Usage

test(
  code,
  kappa_threshold = 0.65,
  baserate_inflation = 0.2,
  ignore_warning = FALSE,
  ...
)

Arguments

code

TBD
kappa_threshold

TBD
baserate_inflation

TBD
ignore_warning

logical, when FALSE (default) verfies test set size before executing test
...

TBD

Value

code object

Unseen Words

Description

Unseen Words

Usage

unseen_words(
  code,
  unweighted = TRUE,
  include_handcoded = FALSE,
  exclude_matched_by = c("excerpts", "word")
)

Arguments

code

Code object
unweighted

logical TRUE (default), binarized the document matrix, so multiple occurrences of single word in one line doesn't count multiple times
include_handcoded

logical FALSE (default), will not use words from the handcoded set. If TRUE, the handcoded set will be used
exclude_matched_by

character, either "excerpts" (default) or "word". "excerpts" will remove full excerpts that match an expression, "word" will remove words from WDM to not include in sum of excerpts

Value

Excerpts that contain unseen words