Function that adds the clustering result to a seqimp object obtained with the seqimpute function

Description

Function that adds the clustering result to a seqimp object obtained with the seqimpute function

Usage

addcluster(impdata, clustering)

Arguments

Value

Returns a seqimp object containing the cluster to which each sequence in each imputed dataset belongs. Specifically, a column named cluster is added to the imputed datasets.

Transform an object of class seqimp into a dataframe or a mids object

Description

The function converts a seqimp object into a specified format.

Usage

fromseqimp(data, format = "long", include = FALSE)

Arguments

Details

The argument format specifies the object that should be returned by the function. It can take the following values

"long"

produces a data set in which imputed data sets are stacked vertically. The following columns are added: 1) .imp referring to the imputation number, and 2) .id the row names of the original dataset

"stacked"

the same as "long", but without the inclusion of the two columns .imp and .id

"mids"

produces an object of class mids, which is the format used by the mice package.

Value

Transform a seqimp object into the desired format.

Author(s)

Kevin Emery

Examples

## Not run: 
# Imputation with the MICT algorithm
imp <- seqimpute(data = gameadd, var = 1:4)

# The object imp is transformed to a dataframe, where completed datasets are
# stacked vertically

imp.stacked <- fromseqimp(
  data = imp,
  format = "stacked", include = FALSE
)

## End(Not run)

Example data set: Game addiction

Description

Dataset containing variables on the gaming addiction of young people. The data consists of gaming addiction, coded as either 'no' or 'yes', measured over four consecutive years for 500 individuals, three covariates and one time-dependent covariate. The yearly states are recorded in columns 1 (T1_abuse) to 4 (T4_abuse).

The three covariates are

• Gender (female or male),

• Age (measured at time 1),

• Track (school or apprenticeship).

The time-varying covariate consists of the individual's relationship to gambling at each of the four time points, appearing in columns T1_gambling, T2_gambling, T3_gambling, and T4_gambling. The states are either no, gambler or problematic gambler

Usage

data(gameadd)

Format

A data frame containing 500 rows, 4 states variable, 3 covariates and a time-dependent covariate.

Plot a seqimp object

Description

Plot a seqimp object. The state distribution plot of the first m completed datasets is shown, possibly alongside the original dataset with missing data

Usage

## S3 method for class 'seqimp'
plot(x, m = 5, include = TRUE, ...)

Arguments

Author(s)

Kevin Emery

Print a seqimp object

Description

Print a seqimp object

Usage

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

Arguments

Author(s)

Kevin Emery

Summary of the types of gaps among a dataset

Description

The seqQuickLook() function aimed at providing an overview of the number and size of the different types of gaps spread in the original dataset.

Usage

seqQuickLook(data, var = NULL, np = 1, nf = 1)

Arguments

Details

The distinction between internal and SLG gaps depends on the number of previous (np) and future (nf) observations that are set for the MICT and MICT-timing algorithms.

Value

Returns a data.frame object that summarizes, for each type of gaps (Internal Gaps, Initial Gaps, Terminal Gaps, LEFT-hand side SLG, RIGHT-hand side SLG, Both-hand side SLG), the minimum length, the maximum length, the total number of gaps and the total number of missing they contain.

Author(s)

Andre Berchtold and Kevin Emery

Examples

data(gameadd)

seqQuickLook(data = gameadd, var = 1:4, np = 1, nf = 1)

Spotting impossible transitions in longitudinal categorical data

Description

The purpose of seqTrans is to spot impossible transitions in longitudinal categorical data.

Usage

seqTrans(data, var = NULL, trans)

Arguments

Value

It returns a matrix where each row is the position of an impossible transition.

Author(s)

Andre Berchtold and Kevin Emery

Examples

data(gameadd)

seqTransList <- seqTrans(data = gameadd, var = 1:4, trans = c("yes->no"))

Generation of missing on longitudinal categorical data.

Description

Generation of missing data in sequence based on a Markovian approach.

Usage

seqaddNA(
  data,
  var = NULL,
  states.high = NULL,
  propdata = 1,
  pstart.high = 0.1,
  pstart.low = 0.005,
  pcont = 0.66,
  maxgap = 3,
  maxprop = 0.75,
  only.traj = FALSE
)

Arguments

Details

The first time point of a trajectory has a pstart.low probability to be missing. For the next time points, the probability to be missing depends on the previous time point. There are four cases:

1. If the previous time point is missing and the maximum length of a missing gap, which is specified by the argument maxgap, is reached, the time point is set as observed.

2. If the previous time point is missing, but the maximum length of a gap is not reached, there is a pcont probability that this time point is missing.

3. If the previous time point is observed and the previous time point belongs to the list of states specified by pstart.high, the probability to be missing is pstart.high.

4. If the previous time point is observed but the previous time point does not belong to the list of states specified by pstart.high, the probability to be missing is pstart.low.

If the proportion of missing data in a given trajectory exceeds the proportion specified by maxprop, the missing data simulation is repeated for the sequence.

Value

A data frame with simulated missing data.

Author(s)

Kevin Emery

Examples

# Generate MCAR missing data on the mvad dataset
# from the TraMineR package

## Not run: 
data(mvad, package = "TraMineR")
mvad.miss <- seqaddNA(mvad, var = 17:86)


# Generate missing data on mvad where joblessness is more likely to trigger
# a missing data gap
mvad.miss2 <- seqaddNA(mvad, var = 17:86, states.high = "joblessness")

## End(Not run)

Extract all the trajectories without missing value.

Description

Extract all the trajectories without missing value.

Usage

seqcomplete(data, var = NULL)

Arguments

Value

Returns either a data frame or a state sequence object, depending the type of data that was provided to the function

Author(s)

Kevin Emery

Examples

# Game addiction dataset
data(gameadd)
# Extract the trajectories without any missing data
gameadd.complete <- seqcomplete(gameadd, var = 1:4)

seqimpute: Imputation of missing data in longitudinal categorical data

Description

The seqimpute package implements the MICT and MICT-timing methods. These are multiple imputation methods for longitudinal data. The core idea of the algorithms is to fills gaps of missing data, which is the typical form of missing data in a longitudinal setting, recursively from their edges. The prediction is based on either a multinomial or a random forest regression model. Covariates and time-dependent covariates can be included in the model.

The MICT-timing algorithm is an extension of the MICT algorithm designed to address a key limitation of the latter: its assumption that position in the trajectory is irrelevant.

Usage

seqimpute(
  data,
  var = NULL,
  np = 1,
  nf = 1,
  m = 5,
  timing = FALSE,
  frame.radius = 0,
  covariates = NULL,
  time.covariates = NULL,
  regr = "multinom",
  npt = 1,
  nfi = 1,
  ParExec = FALSE,
  ncores = NULL,
  SetRNGSeed = FALSE,
  end.impute = TRUE,
  verbose = TRUE,
  available = TRUE,
  pastDistrib = FALSE,
  futureDistrib = FALSE,
  ...
)

Arguments

Details

The imputation process is divided into several steps, depending on the type of gaps of missing data. The order of imputation of the gaps are:

Internal gap:

there is at least np observations before an internal gap and nf after the gap

Initial gap:

gaps situated at the very beginning of a trajectory

Terminal gap:

gaps situated at the very end of a trajectory

Left-hand side specifically located gap (SLG):

gaps that have at least nf observations after the gap, but less than np observation before it

Right-hand side SLG:

gaps that have at least np observations before the gap, but less than nf observation after it

Both-hand side SLG:

gaps that have less than np observations before the gap, and less than nf observations after it

The primary difference between the MICT and MICT-timing algorithms lies in their approach to selecting patterns from other sequences for fitting the multinomial model. While the MICT algorithm considers all similar patterns regardless of their temporal placement, MICT-timing restricts pattern selection to those that are temporally closest to the missing value. This refinement ensures that the imputation process adequately accounts for temporal dynamics, imping in more accurate imputed values.

Value

An object of class seqimp, which is a list with the following elements:

data

A data.frame containing the original (incomplete) data.

imp

A list of m data.frame corresponding to the imputed datasets.

m

The number of imputations.

method

A character vector specifying whether MICT or MICT-timing was used.

np

Number of prior states included in the imputation model.

nf

Number of subsequent states included in the imputation model.

regr

A character vector specifying whether multinomial or random forest imputation models were applied.

call

The call that created the object.

Author(s)

Kevin Emery <kevin.emery@unige.ch>, Andre Berchtold, Anthony Guinchard, and Kamyar Taher

References

Halpin, B. (2012). Multiple imputation for life-course sequence data. Working Paper WP2012-01, Department of Sociology, University of Limerick. http://hdl.handle.net/10344/3639.

Halpin, B. (2013). Imputing sequence data: Extensions to initial and terminal gaps, Stata's. Working Paper WP2013-01, Department of Sociology, University of Limerick. http://hdl.handle.net/10344/3620

Emery, K., Studer, M., & Berchtold, A. (2024). Comparison of imputation methods for univariate categorical longitudinal data. Quality & Quantity, 1-25. https://link.springer.com/article/10.1007/s11135-024-02028-z

Examples

# Default multiple imputation of the trajectories of game addiction with the
# MICT algorithm

## Not run: 
set.seed(5)
imp1 <- seqimpute(data = gameadd, var = 1:4)


# Default multiple imputation with the MICT-timing algorithm
set.seed(3)
imp2 <- seqimpute(data = gameadd, var = 1:4, timing = TRUE)


# Inclusion in the MICt-timing imputation process of the three background
# characteristics (Gender, Age and Track), and the time-varying covariate
# about gambling


set.seed(4)
imp3 <- seqimpute(
  data = gameadd, var = 1:4, covariates = 5:7,
  time.covariates = 8:11
)


# Parallel computation


imp4 <- seqimpute(
  data = gameadd, var = 1:4, covariates = 5:7,
  time.covariates = 8:11, ParExec = TRUE, ncores = 5, SetRNGSeed = 2
)

## End(Not run)

Plot all the patterns of missing data.

Description

This function plots all patterns of missing data within sequences, based on the seqIplot function.

Usage

seqmissIplot(data, var = NULL, with.complete = TRUE, void.miss = TRUE, ...)

Arguments

Details

This function uses seqIplot to visualize all patterns of missing data within sequences. For further customization options, refer to the seqIplot documentation.

Author(s)

Kevin Emery

Examples

# Plot all the patterns of missing data

seqmissIplot(gameadd, var = 1:4)

# Plot all the patterns of missing data discarding
# complete trajectories

seqmissIplot(gameadd, var = 1:4, with.missing = FALSE)

Plot the most common patterns of missing data.

Description

This function plots the most frequent patterns of missing data, based on the seqfplot function.

Usage

seqmissfplot(data, var = NULL, with.complete = TRUE, void.miss = TRUE, ...)

Arguments

Details

This plot function is based on the seqfplot function, allowing users to visualize patterns of missing data within sequences. For details on additional customizable arguments, see the seqfplot documentation.

By default, this function plots the 10 most frequent patterns. The number of patterns to be plotted can be adjusted using the idxs argument in seqfplot.

Author(s)

Kevin Emery

Examples

# Plot the 10 most common patterns of missing data

seqmissfplot(gameadd, var = 1:4)

# Plot the 10 most common patterns of missing data discarding
# complete trajectories

seqmissfplot(gameadd, var = 1:4, with.missing = FALSE)

# Plot only the 5 most common patterns of missing data discarding
# complete trajectories

seqmissfplot(gameadd, var = 1:4, with.missing = FALSE, idxs = 1:5)

Identification and visualization of states that best characterize sequences with missing data

Description

This function identifies and visualizes states that best characterize sequences with missing data at each position (time point), comparing them to sequences without missing data at each position (time point). It is based on the seqimplic function. For more information on the methodology, see the seqimplic documentation.

Usage

seqmissimplic(data, var = NULL, void.miss = TRUE, ...)

Arguments

Value

returns a seqimplic object that can be plotted and printed.

Author(s)

Kevin Emery

Examples

# For illustration purpose, we simulate missing data on the mvad dataset,
# available in the TraMineR package. The state "joblessness" state has a
# higher probability of triggering a missing gap

## Not run: 
data(mvad, package = "TraMineR")
mvad.miss <- seqaddNA(mvad, var = 17:86, states.high = "joblessness")

# The states that best characterize sequences with missing data
implic <- seqmissimplic(mvad.miss, var = 17:86)

# Visualization of the results
plot(implic)

## End(Not run)

Extract all the trajectories with at least one missing value

Description

Extract all the trajectories with at least one missing value

Usage

seqwithmiss(data, var = NULL)

Arguments

Value

Returns either a data frame or a state sequence object, depending the type of data that was provided to the function

Author(s)

Kevin Emery

Examples

# Game addiction dataset
data(gameadd)
# Extract the trajectories without any missing data
gameadd.withmiss <- seqwithmiss(gameadd, var = 1:4)

Summary of a seqimp object

Description

Summary of a seqimp object

Usage

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

Arguments

Author(s)

Kevin Emery