Title:	Timing, Anatomical, Therapeutic and Chemical Based Medication Clustering
Version:	0.2.0
Description:	Agglomerative hierarchical clustering with a bespoke distance measure based on medication similarities in the Anatomical Therapeutic Chemical Classification System, medication timing and medication amount or dosage. Tools for summarizing, illustrating and manipulating the cluster objects are also available.
License:	GPL (≥ 3) | file LICENSE
Encoding:	UTF-8
RoxygenNote:	7.3.2
Imports:	dplyr (≥ 1.0.0), fuzzyjoin, magrittr, purrr (≥ 1.0.0), Rfast, rlang (≥ 1.0.0), stats, stringr, tibble (≥ 3.0.0), tidyr (≥ 1.2.0), tidyselect, Rcpp (≥ 1.0.8), ggplot2 (≥ 3.3.0), scales, utils
Suggests:	rmarkdown, knitr, testthat (≥ 3.0.0)
BugReports:	https://github.com/Laksafoss/tame/issues
Config/testthat/edition:	3
Depends:	R (≥ 4.2)
LazyData:	true
LinkingTo:	Rcpp
NeedsCompilation:	yes
Packaged:	2025-04-02 14:03:17 UTC; B245807
Author:	Anna Laksafoss [aut, cre]
Maintainer:	Anna Laksafoss <adls@ssi.dk>
Repository:	CRAN
Date/Publication:	2025-04-02 14:20:02 UTC

Timing and ATC based Medication Clustering

Description

Agglomerative hierarchical clustering with a bespoke distance measure based on medication ATC codes similarities, medication timing and medication amount or dosage. Tools for summarizing, illustrating and manipulating the cluster objects are also available.

Author(s)

Maintainer: Anna Laksafoss adls@ssi.dk (ORCID)

See Also

Useful links:

• Report bugs at https://github.com/Laksafoss/tame/issues

The Frequency of Assignment to Each Cluster

Description

The function cluster_frequency() calculates the number and frequency of individuals assigned to each cluster.

Usage

cluster_frequency(
  object,
  only = NULL,
  clusters = NULL,
  additional_data = NULL,
  ...
)

Arguments

Details

cluster_frequency() calculates the number of individuals assigned to each cluster and the associated frequency of assignment.

Value

cluster_frequency() returns a data frame with class cluster_frequency.

• Clustering the name of the clustering.

• Cluster the cluster name.

• Count the number of individuals assigned to the cluster.

• Percent the percent of individuals assigned to the cluster.

Examples

clust <- medic(complications, id = id, atc = atc, k = 3:5)

# make frequency tables
cluster_frequency(clust, k == 5)
cluster_frequency(clust, k < 5, I:III)

Finding chosen cluster names

Description

The cluster_selector() function finds the name of the chosen clusters.

Usage

cluster_selector(clustering, clusters = NULL)

Arguments

Details

The cluster_selector() finds the names of the chosen clusters present in the clustering medic object. This function is used in all of the investigative functions called by summary().

Value

A character vector with the names of the chosen clusters.

See Also

method_selector() is another selector method used through out the package.

enrich() is a joining function used for enriching the clustering characteristics with user-defined data.

summary.medic() is the main function for summarizing the clustering results.

Frequency tables for medication amount

Description

The function comedication_count() calculates the number of unique medications for each individual and presents the count frequencies by cluster.

Usage

comedication_count(
  object,
  only = NULL,
  clusters = NULL,
  count_grouper = function(x) {
     cut(x, breaks = c(0, 1, 2, Inf), labels = c("1",
    "2", "3+"))
 },
  additional_data = NULL,
  ...
)

Arguments

Details

comedication_count() calculates the number of ATC codes an individual has, and then outputs the number of individuals within a cluster that has that many ATC codes. Moreover, various relevant percentages or calculated. See Value below for more details on these percentages.

Value

comedication_count() returns a data frame of class comedication_count

• Clustering the name of the clustering.

• Cluster the name of the cluster.

• ⁠Medication Count⁠ a number of medications. The numbers or groups are given by the count_grouper() function.

• ⁠Number of People⁠ the number of individuals in cluster who has ⁠Medication Count⁠ number of comedications in study.

• ⁠Number of medications⁠ the number of medications of individuals who has ⁠Medication Count⁠ number of comedications in the cluster.

• ⁠Percentage of All People⁠ the percentage of individuals is study who has ⁠Medication Count⁠ number of comedications in the cluster.

• ⁠Percentage of People in Cluster⁠ the percentage of individuals in the cluster who has ⁠Medication Count⁠ number of comedications.

• ⁠Percentage of All Medications⁠ the percentage of medication in study from individuals who has ⁠Medication Count⁠ number of comedications in cluster.

• ⁠Percentage of Medication in Cluster⁠ the percentage of medication in cluster from individuals who has ⁠Medication Count⁠ number of comedications.

• ⁠Percentage of People with the Same Medication Count⁠ percentage of individuals among those with ⁠Medication Count⁠ number of comedications in this cluster.

• ⁠Percentage of Medication with the Same Medication Count⁠ percentage of medication among medication of individuals with ⁠Medication Count⁠ number of comedications in this cluster.

Examples

clust <- medic(complications, id = id, atc = atc, k = 3:5)

comedication_count(clust, k == 5, clusters = I:III)

A Simulated Data Set About Pregnancy Complications

Description

We use this data set in all the examples in the package.

Usage

complications

Format

An object of class data.frame with 149 rows and 8 columns.

Construct color scales for summary plot

Description

The construct_color_scales() function constructs the color scales for the summary plot.

Usage

construct_color_scales(
  plot_data,
  comedication_count_colors = NULL,
  medication_frequency_colors = NULL,
  timing_atc_group_colors = NULL
)

Arguments

Value

A list with the color scales:

• comedication_count_colors a character vector with the colors for the comedication count.

• medication_frequency_colors a character vector with the colors for the medication frequency.

• medication_frequency_linetype a numeric vector with the linetypes for the medication frequency.

• timing_atc_group_colors a character vector with the colors for the ATC groups.

Construct plot data for summary plot

Description

The construct_plot_data() function constructs the data for the summary plot.

Usage

construct_plot_data(
  object,
  time_scale,
  plot_individual,
  labels,
  alpha_individual = 0.1,
  label_y_value = 0,
  ...
)

Arguments

Value

A data frame with the plot data.

Construct time scale for summary plot

Description

The construct_time_scale() function constructs a time scale for the summary plot.

Usage

construct_time_scale(object, n_breaks = 5)

Arguments

Value

A list with the following components:

• translator a data frame with the time scale.

• mid a numeric value specifying the mid point of the time scale.

• width a numeric value specifying the width of the time scale.

• breaks a character vector with the breaks in the time scale.

Default ATC groups for summaries

Description

This function finds the default ATC groups for the summaries. It is used in the summary.medic function.

Usage

default_atc_groups(object, min_n = 2)

Arguments

Value

A data frame with two columns: regex and atc_groups.

A Simulated Data Set About Eczema

Description

A Simulated Data Set About Eczema

Usage

eczema

Format

An object of class data.frame with 50644 rows and 7 columns.

Employ a Clustering to New Data

Description

Employ a clustering to new data

Usage

employ(
  object,
  new_data,
  only = NULL,
  additional_data = NULL,
  assignment_method = "nearest_cluster",
  parallel = FALSE,
  ...
)

Arguments

Value

employ returns a medic object.

Examples

part1 <- complications[1:100,]
part2 <- complications[101:149,]

clust <- medic(part1, id = id, atc = atc, k = 3)

# Nearest cluster matching
employ(clust, part2)

# Only exact matching
employ(clust, part2, assignment_method = "exact_only")

Enrich Clustering Parameter

Description

Enrich the parameter information in a clustering with user-defined data.

Usage

enrich(object, additional_data = NULL, by = NULL)

Arguments

Details

The enrich() function is a joining function used for enriching the clustering characteristics with user-defined data. This function is used in all of the investigative functions with a additional_data statement such as summary(), cluster_frequency() and medication_frequency().

Value

An object of class medic.

Examples

clust <- medic(
   complications,
   id = id,
   atc = atc,
   timing = first_trimester:third_trimester,
   k = 3:5
)

new_parameters <- data.frame(k = 3:5, size = c("small", "small", "large"))

enrich(clust, new_parameters)

Flatten trajectory data

Description

The flatten_trajctories() function flattens the individual and average trajectories.

Usage

flatten_trajctories(object, individual)

Arguments

Value

A data frame with the flattened trajectories.

Test if an object is a medic-object

Description

Test if an object is a medic-object

Usage

is.medic(object)

Arguments

Value

TRUE is the object inherits from the medic class and has the required elements.

Examples

clust <- medic(complications, id = id, atc = atc, k = 3)
is.medic(clust)

Medication clustering (based on ATC and timing)

Description

The medic method uses agglomerative hierarchical clustering with a bespoke distance measure based on medication ATC codes similarities, medication timing and medication amount or dosage.

Usage

medic(
  data,
  k = 5,
  id,
  atc,
  timing,
  base_clustering,
  linkage = "complete",
  summation_method = "sum_of_minima",
  alpha = 1,
  beta = 1,
  gamma = 1,
  p = 1,
  theta = (5:0)/5,
  parallel = FALSE,
  return_distance_matrix = FALSE,
  set_seed = FALSE,
  ...
)

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

Arguments

Details

The medic method uses agglomerative hierarchical clustering with a bespoke distance measure based on medication ATC codes and timing similarities to assign medication pattern clusters to people.

Two versions of the distance measure are available:

The double sum:

% d(p_i, p_j) = N_{\alpha}(M_i \times M_j) \sum_{m\in M_i}\sum_{n \in M_j}% ((1 + D_{\theta}(m,n)) (1 + \gamma T_p(t_{im},t_{jn})) - 1)^{\beta}.%

and the sum of minima:

% d(p_i, p_j) = \frac{1}{2}(N_{\alpha}(M_i)\sum_{m\in M_i}\min_{n \in M_j}% ((1 + D_{\theta}(m,n)) (1 + \gamma T_p(t_{im},t_{jn})) - 1)^{\beta} + N_{\alpha}(M_j) \sum_{n\in M_j}\min_{m \in M_i}% ((1 + D_{\theta}(m,n)) (1 + \gamma T_p(t_{im},t_{jn})) - 1)^{\beta}).%

Normalization

% N_{\alpha}(x) = |x|^{-\alpha}%

If the normalization tuning, alpha, is 0, then no normalization is preformed and the distance measure becomes highly dependent on the number of distinct medications given. That is, people using more medication will have larger distances to others. If the normalization tuning, alpha, is 1 - the default - then the summation is normalized with the number of terms in the sum, in other words, the average is calculated.

ATC distance

The central idea of this method, namely the ATC distance, is given as

% D_{\theta}(x, y) = \sum_{i=1,...,5}1\{x and y match on level i, but not level i + 1\}\theta_i%

The ATC distance is tuned using the vector theta.

Note that two ATC codes are said to match at level i when they are identical at level i. E.g. the two codes N06AB01 and N06AA01 match on level 1, 2, and 3 as they are both "N" at level 1, "N06" at level 2, and "N06A" at level 3, but at level 4 they differ ("N06AB" and "N06AA" are not the same).

Timing distance

The timing distance is a simple Minkowski distance:

% T(x,y) =(\sum_{t \in T} |x_t - y_t|^p)^{1/p}.%

When p is 1, the default, the Manhattan distance is used.

Value

An object of class medic which describes the clusters produced the hierarchical clustering process. The object is a list with components:

data

the inputted data frame data with the cluster assignments appended at the end.

clustering

a data frame with the person id as given by id, the .analysis_order and the clusters found.

variables

a list of the variables used in the clustering.

parameters

a data frame with all the inputted clustering parameters and the corresponding method names. These method names correspond to the column names for each cluster in the clustering data frame described right above.

key

a list of keys used internally in the function to keep track of simplified versions of the data.

distance_matrix

the distance matrices for each method if return_distance_matrix is TRUE otherwise NULL.

call

the matched call.

Methods (by generic)

• print(medic): Print method for medic-objects

See Also

summary.medic for summaries and plots.

employ for employing an existing clustering to new data.

enrich for enriching the meta data in the medic object with additional data.

Examples

# A simple clustering based only on ATC
clust <- medic(complications, id = id, atc = atc, k = 3)

# A simple clustering with both ATC and timing
clust <- medic(
  complications,
  id = id,
  atc = atc,
  timing = first_trimester:third_trimester,
  k = 3
)

ATC Code Frequency Within Clusters

Description

The function medications() calculates the frequency of the different unique ATC codes within each cluster.

Usage

medication_frequency(
  object,
  only = NULL,
  clusters = NULL,
  additional_data = NULL,
  ...
)

Arguments

Details

medication_frequency() calculates the number of individuals with a specific ATC code within a cluster. Moreover, it calculates the percentage of people with this medication assigned to this cluster and the percent of people within the cluster with this medication.

Value

medication_frequency() returns a data frame with class medication_frequency.

• Clustering the name of the clustering.

• Cluster the cluster name.

• atc ATC codes.

• Count number of individuals with this ATC code in this cluster.

• ⁠Percent of All Medication⁠ the percentage of individuals in the study with this ATC code and cluster.

• ⁠Percent of Medication in Cluster⁠ the percent of individuals in the cluster with this ATC code.

Examples

clust <- medic(complications, id = id, atc = atc, k = 3:5)

medication_frequency(clust, k == 5, clusters = I:III)

Method Selector

Description

The method_selector() is a filtering function for extracting only the relevant parameters from a medic object.

Usage

method_selector(clustering, only, additional_data = NULL)

Arguments

Details

The method_selector() function is a filtering function used for extracting the characteristics of the chosen method. This function is used in all of the investigative functions called by summary().

Value

A parameters data frame with at least the following columns

• cluster_name the name of the clustering cluster_name.

• clustering the name of the clustering group before the number of clusters is chosen.

• linkage the linkage criteria.

• alpha the normalization tuning.

• beta the timing importance tuning.

• gamma the dose importance tuning.

• theta the ATC measure tuning.

• k the number of clusters.

where each row specifies the parameters for the chosen methods. Any additional columns present in the the outcome are from the additional_data.

See Also

cluster_selector() is another selector method used through out the package.

enrich() is a joining function used for enriching the clustering characteristics with user-defined data.

summary.medic() is the main function for summarizing the clustering results.

Internal option constructor

Description

Given the input of the medic this function checks the input and constructs a data frame with the analysis parameters specified by the user.

Usage

parameters_constructor(
  data,
  id,
  k = 5,
  atc,
  timing,
  base_clustering,
  linkage = "complete",
  summation_method = "sum_of_minima",
  alpha = 1,
  beta = 1,
  gamma = 1,
  p = 1,
  theta = (5:0)/5,
  ...
)

Arguments

Value

A data.frame with the parameters for clustering.

Examples

parameters_constructor(
   data = complications,
   k = 3,
   id = id,
   atc = atc
)

Plot Cluster Frequency

Description

This function plots the cluster frequency.

Usage

plot_cluster_frequency(object, ...)

## S3 method for class 'medic'
plot_cluster_frequency(object, ...)

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

## S3 method for class 'cluster_frequency'
plot_cluster_frequency(object, scale = "percent", with_population = FALSE, ...)

Arguments

Value

A ggplot object.

See Also

cluster_frequency

plot_medication_frequency

plot_comedication_count

plot_timing_trajectory

plot_timing_atc_group

plot_summary

Examples

clust <- medic(complications, id = id, atc = atc, k = 3)
clust |> plot_cluster_frequency()
clust |> cluster_frequency() |> plot_cluster_frequency()
clust |> summary() |> plot_cluster_frequency()

Plot Comedication Count

Description

This function plots the comedication count.

Usage

plot_comedication_count(object, ...)

## S3 method for class 'medic'
plot_comedication_count(object, ...)

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

## S3 method for class 'comedication_count'
plot_comedication_count(
  object,
  scale = "percent",
  scope = "cluster",
  focus = "people",
  with_population = FALSE,
  ...
)

Arguments

Value

A ggplot object.

See Also

comedication_count

plot_cluster_frequency

plot_medication_frequency

plot_timing_trajectory

plot_timing_atc_group

plot_summary

Examples

clust <- medic(complications, id = id, atc = atc, k = 3)

clust |> plot_comedication_count()
clust |> comedication_count() |> plot_comedication_count()
clust |> summary() |> plot_comedication_count()

Plot Medication Frequency

Description

This function plots the medication frequency.

Usage

plot_medication_frequency(object, ...)

## S3 method for class 'medic'
plot_medication_frequency(object, ...)

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

## S3 method for class 'medication_frequency'
plot_medication_frequency(
  object,
  scale = "percent",
  scope = "cluster",
  with_population = FALSE,
  ...
)

Arguments

Value

A ggplot object.

See Also

medication_frequency

plot_cluster_frequency

plot_comedication_count

plot_timing_trajectory

plot_timing_atc_group

plot_summary

Examples

clust <- medic(complications, id = id, atc = atc, k = 3)

clust |> plot_medication_frequency()
clust |> medication_frequency() |> plot_medication_frequency()
clust |> summary() |> plot_medication_frequency()

Plot Summary

Description

This function plots the summary of the clustering results.

Usage

plot_summary(object, ...)

## S3 method for class 'medic'
plot_summary(object, only = NULL, clusters = NULL, additional_data = NULL, ...)

## S3 method for class 'summary.medic'
plot_summary(
  object,
  n_breaks = 5,
  plot_individual = FALSE,
  labels = FALSE,
  alpha_individual = 0.1,
  label_y_value = 0.1,
  ...
)

Arguments

Value

A ggplot object.

See Also

summary

plot_cluster_frequency

plot_medication_frequency

plot_comedication_count

plot_timing_trajectory

plot_timing_atc_group

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3,
  timing = first_trimester:third_trimester
)

clust |> plot_summary()
clust |> summary() |> plot_summary()

# If the clustering object contains more than one clustering, it is necessary
# to filter the clustering, as only one clustering can be plotted at a time.
clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)
clust |> plot_summary(only = k == 4)
clust |> summary(only = k == 4) |> plot_summary()

Plot Timing ATC Group

Description

This function plots the timing ATC group.

Usage

plot_timing_atc_group(object, ...)

## S3 method for class 'medic'
plot_timing_atc_group(object, ...)

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

## S3 method for class 'timing_atc_group'
plot_timing_atc_group(
  object,
  focus = "average",
  with_population = FALSE,
  max_lines = 50,
  ...
)

Arguments

Value

A ggplot object.

See Also

timing_atc_group

plot_cluster_frequency

plot_medication_frequency

plot_comedication_count

plot_timing_trajectory

plot_summary

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)

clust |> plot_timing_atc_group()
clust |> timing_atc_group() |> plot_timing_atc_group()
clust |> summary() |> plot_timing_atc_group()

Plot Timing Trajectory

Description

This function plots the timing trajectory.

Usage

plot_timing_trajectory(object, ...)

## S3 method for class 'medic'
plot_timing_trajectory(object, ...)

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

## S3 method for class 'timing_trajectory'
plot_timing_trajectory(
  object,
  focus = "average",
  with_population = FALSE,
  max_lines = 50,
  ...
)

Arguments

Value

A ggplot object.

See Also

timing_trajectory

plot_cluster_frequency

plot_medication_frequency

plot_comedication_count

plot_timing_atc_group

plot_summary

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)

clust |> plot_timing_trajectory()
clust |> timing_trajectory() |> plot_timing_trajectory()
clust |> summary() |> plot_timing_trajectory()

Print Summary of Medication

Description

This function prints a summary of medication information.

Usage

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

Arguments

Details

This function prints various information about medication, including cluster frequency, medication frequency, number of different medication taken in the study period, average exposure trajectories, and average exposure trajectories by ATC groups.

Value

The function is called for its side effects and does not return any value.

Refactor Cluster Levels

Description

Refactor the levels of the chosen clusters.

Usage

refactor(object, ..., inheret_parameters = TRUE)

Arguments

Value

A medic object with relevant clusterings refactored.

Examples

clust <- medic(complications, id = id, atc = atc, k = 3:4)

# Refactor one clustering
refactor(
  clust,
  `cluster_1_k=4` = dplyr::recode(`cluster_1_k=4`, IV = "III")
)

# Refactor all clusterings
refactor(
  clust,
  dplyr::across(
    dplyr::everything(),
    ~dplyr::recode(., IV = "III")
  )
)

Summary of a medic-object using str function

Description

Summary of a medic-object using str function

Usage

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

Arguments

Summary of medic object

Description

Make cluster characterizing summaries.

Usage

## S3 method for class 'medic'
summary(
  object,
  only = NULL,
  clusters = NULL,
  outputs = "all",
  additional_data = NULL,
  ...
)

Arguments

Value

A list of clustering characteristics of class summary.medic is returned. It can contain any of the following characteristics:

Cluster Frequencies

The number of individuals assigned to each cluster and the associated frequency of assignment.

Medication Frequencies

The number of individuals with a specific ATC code within a cluster. Moreover, it calculates the percentage of people with this medication assigned to this cluster and the percent of people within the cluster with this medication.

Comedication Count

The number of ATC codes an individual has, and then outputs the number of individuals within a cluster that has that many ATC codes. Moreover, various relevant percentages or calculated. See Value below for more details on these percentages.

Timing Trajectories

The number of unique timing trajectories in each cluster, and the average timing trajectories in each cluster.

Timing and ATC group interactions

The number of people with unique timing trajectory and ATC group, as given by atc_groups, in each cluster.

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)
summary(clust)

Crop Clustering Summary

Description

Functions for cropping summarized cluster data.

Usage

summary_crop(object, ...)

## S3 method for class 'cluster_frequency'
summary_crop(object, top_n = 5L, min_count = 0, min_percent = 0, ...)

## S3 method for class 'medication_frequency'
summary_crop(
  object,
  top_n = 5L,
  min_count = 0,
  min_percent = 0,
  scope = "cluster",
  ...
)

## S3 method for class 'comedication_count'
summary_crop(object, ...)

## S3 method for class 'timing_trajectory'
summary_crop(object, sample_n_individual = 100L, weighted_sample = TRUE, ...)

## S3 method for class 'timing_atc_group'
summary_crop(
  object,
  sample_n_individual = 100L,
  weighted_sample = TRUE,
  min_count = 0L,
  ...
)

## S3 method for class 'summary.medic'
summary_crop(object, which = "all", ...)

Arguments

Value

A summary object, which is a modified version of the input summary object.

cluster_frequency summary crop

Extracts the top top_n clusters by count. If top_n is Inf, all clusters are kept. If min_count is greater than 0, clusters with a count less than min_count are removed. If min_percent is greater than 0, clusters with a percentage less than min_percent are removed. The remaining clusters are grouped into a "Remaining" cluster.

medication_frequency summary crop

Extracts the top top_n medications by count. If top_n is Inf, all medications are kept. If min_count is greater than 0, medications with a count less than min_count are removed. If min_percent is greater than 0, medications with a percentage less than min_percent are removed. The remaining medications are grouped into a "Remaining" cluster.

The scope argument determines the scope of the crop. If scope is "cluster", the crop is based on the percentage of medication in the cluster. If scope is "global", the crop is based on the percentage of all medication.

comedication_count summary crop

TO DO

timing_trajectory summary crop

Samples sample_n_individual individual timing trajectories. If sample_n_individual is Inf, all individual timing trajectories are kept. If weighted_sample is TRUE, the individual timing trajectories are sampled weighted by the number of medications in the individual timing trajectory.

timing_atc_group summary crop

Samples sample_n_individual individual timing trajectories. If sample_n_individual is Inf, all individual timing trajectories are kept. If weighted_sample is TRUE, the individual timing trajectories are sampled weighted by the number of medications in the individual timing trajectory.

summary.medic summary crop

Crops multiple summaries. The which argument is a character vector specifying which summaries to crop. The options are "cluster_frequency", "medication_frequency", "comedication_count", "timing_trajectory", and "timing_atc_group". If which is "all", all summaries are cropped.

The ... argument is passed to the specific methods, e.g. top_n and min_count are passed to cluster_frequency and medication_frequency.

See Also

summary, cluster_frequency, medication_frequency, comedication_count, timing_trajectory, timing_atc_group

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)


# Crop the cluster frequency summary
clust |>
  cluster_frequency() |>
  summary_crop(top_n = 3)

clust |>
  summary() |>
  summary_crop(which = "cluster_frequency", top_n = 3)


# Crop the medication frequency summary
clust |>
  medication_frequency() |>
  summary_crop(top_n = 3)

clust |>
  summary() |>
  summary_crop(which = "medication_frequency", top_n = 3)


# Crop the co-medication count summary
clust |>
  comedication_count() |>
  summary_crop(min_count = 10)

clust |>
  summary() |>
  summary_crop(which = "comedication_count", min_count = 10)


# crop the timing trajectory summary
clust |>
  timing_trajectory() |>
  summary_crop()

clust |>
  summary() |>
  summary_crop(which = "timing_trajectory")


# crop the timing ATC group summary
clust |>
  timing_atc_group() |>
  summary_crop()

clust |>
  summary() |>
  summary_crop(which = "timing_atc_group")

# crop multiple summaries
clust |>
  summary() |>
  summary_crop(
    which = c("cluster_frequency", "medication_frequency"),
    top_n = 3
  )

Timing and ATC pattern interactions

Description

The function timing_atc_group() calculates the frequencies of distinct timing and ATC combinations within clusters.

Usage

timing_atc_group(
  object,
  only = NULL,
  clusters = NULL,
  atc_groups = default_atc_groups,
  additional_data = NULL,
  ...
)

Arguments

Details

timing_atc_group() calculates both the number of people with unique timing trajectory and ATC group, as given by atc_groups, in each cluster.

Value

timing_atc_group() returns a list of class timing_atc_group with two data frames:

average

• Clustering the name of the clustering.

• Cluster the name of the cluster.

• ⁠ATC Groups⁠ the name of the ATC group. The groups are given by the atc_groups input.

• timing variables the average timing value in the ATC group and cluster.

• ⁠Number of Medications⁠ the number of medications in the ATC group in the cluster.

• ⁠Percentage of Medications⁠ the percentage of medication in the cluster with this ATC group.

• ⁠Number of Distinct Timing Trajectories⁠ the number of unique timing trajectories in the ATC group in the cluster.

individual

• Clustering the name of the clustering.

• Cluster the name of the cluster.

• timing variables a unique timing pattern in the ATC group and cluster.

• ⁠Number of Medications with Timing Trajectory⁠ the number of medications with this unique timing trajectory and ATC group.

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)

timing_atc_group(clust, k == 5, clusters = I:III)

Timing pattern frequency within clusters

Description

timing_trajectory() calculates the average timing paths within clusters.

Usage

timing_trajectory(
  object,
  only = NULL,
  clusters = NULL,
  additional_data = NULL,
  ...
)

Arguments

Details

timing_trajectory() calculates both the number of unique timing trajectories in each cluster and the average timing trajectories in each cluster.

Value

timing_trajectory() returns a list of class timing_trajectory with two data frames:

average

• Clustering the name of the clustering.

• Cluster the cluster name.

• timing variables the average timing value in the cluster.

• Count the number of people in the cluster.

individual

• Clustering the name of the clustering.

• Cluster the cluster name.

• timing variables unique timing pattern in the cluster.

• Count number of people with this unique timing pattern.

Examples

clust <- medic(
  complications,
  id = id,
  atc = atc,
  k = 3:5,
  timing = first_trimester:third_trimester
)

timing_trajectory(clust, k == 5, clusters = I:III)