Sub-criteria attributes.

Description

Recursive evaluation of a function (func) on each attribute (vector) in a sub_criteria.

Usage

attr_eval(x, func = length, simplify = TRUE)

Arguments

Value

vector; list

Examples

x <- sub_criteria(rep(1, 5), rep(5 * 10, 5))
attr_eval(x)
attr_eval(x, func = max)
attr_eval(x, func = max, simplify = FALSE)
attr_eval(sub_criteria(x, x), func = max, simplify = FALSE)

Vectorised approach to group operations.

Description

Vectorised approach to group operations.

Usage

bys_count(by)

bys_rank(..., by = NULL, from_last = FALSE)

bys_position(val, by = NULL, from_last = FALSE, ordered = TRUE)

bys_val(..., val, by = NULL, from_last = FALSE)

bys_nval(..., val, by = NULL, from_last = FALSE, n = 1, nmax = FALSE)

bys_min(val, by = NULL, na.rm = TRUE)

bys_max(val, by = NULL, na.rm = TRUE)

bys_sum(val, by = NULL, na.rm = TRUE)

bys_prod(val, by = NULL, na.rm = TRUE)

bys_cummin(val, by = NULL, na.rm = TRUE)

bys_cummax(val, by = NULL, na.rm = FALSE)

bys_cumsum(val, by = NULL, na.rm = TRUE)

bys_cumprod(val, by = NULL, na.rm = TRUE)

bys_lag(val, by = NULL, n = 1)

bys_lead(val, by = NULL, n = 1)

Arguments

Value

[atomic]

Examples

x <- data.frame(
  group = c(2, 2, 1, 2, 1, 1, 1, 2, 1, 1),
  value = c(13, 14, 20, 9, 2, 1, 8, 18, 3, 17))

bys_count(x$group)
bys_position(x$value, by = x$group, from_last = TRUE)
bys_rank(by = x$group, val = x$value, from_last = TRUE)
bys_val(x$value, by = x$group, val = x$value, from_last = TRUE)
bys_nval(x$value, by = x$group, val = x$value, from_last = TRUE, n = 2)
bys_min(by = x$group, val = x$value)
bys_max(by = x$group, val = x$value)
bys_sum(by = x$group, val = x$value)
bys_prod(by = x$group, val = x$value)
bys_cummin(by = x$group, val = x$value)
bys_cummax(by = x$group, val = x$value)
bys_cumsum(by = x$group, val = x$value)
bys_cumprod(by = x$group, val = x$value)
bys_lag(by = x$group, val = x$value)
bys_lead(by = x$group, val = x$value)

Vector combinations

Description

Numeric codes for unique combination of vectors.

Usage

combi(...)

Arguments

Value

numeric

Examples

x <- c("A", "B", "A", "C", "B", "B")
y <- c("X", "X", "Z", "Z", "X", "Z")
combi(x, y)

# The code above is equivalent to but quicker than the one below.
z <- paste0(y, "-", x)
z <- match(z, z)
z

Nested sorting

Description

Returns a sort order after sorting by a vector within another vector.

Usage

custom_sort(..., decreasing = FALSE, unique = FALSE)

Arguments

Value

numeric sort order.

Examples

a <- c(1, 1, 1, 2, 2)
b <- c(2, 3, 2, 1, 1)

custom_sort(a, b)
custom_sort(b, a)
custom_sort(b, a, unique = TRUE)

d_report

Description

d_report

Usage

## S3 method for class 'd_report'
plot(
  x,
  ...,
  metric = c("cumulative_duration", "duration", "max_memory", "records_checked",
    "records_skipped", "records_assigned")
)

## S3 method for class 'd_report'
as.list(x, ...)

## S3 method for class 'd_report'
as.data.frame(x, ...)

Arguments

Unlink group identifiers

Description

Unlink records from an episode (epid), record group (pid) or pane (pane) object.

Usage

delink(x, lgk, ...)

## S3 method for class 'epid'
delink(x, lgk, ...)

## S3 method for class 'pane'
delink(x, lgk, ...)

## S3 method for class 'pid'
delink(x, lgk, ...)

Arguments

Value

epid; pid; pane

Examples

ep <- episodes(1:8)
unlinked_ep <- delink(ep, ep@sn %in% c(3, 8))
ep; unlinked_ep

pn <- partitions(1:8, length.out = 2, separate = TRUE)
unlinked_pn <- delink(pn, pn@.Data == 5)
pn; unlinked_pn

pd <- links(list(c(1, 1, 1, NA, NA),
                 c(NA, NA, 2, 2, 2)))
unlinked_pd <- delink(pd, pd@pid_cri == 1)
pd; unlinked_pd

# A warning is given if an index record is unlinked as this will lead to seemly impossible links.
ep2 <- episodes(1:8, 2, episode_type = "rolling")
unlinked_ep2 <- delink(ep2, ep2@sn %in% c(3, 5))
schema(ep2, custom_label = decode(ep2@case_nm), seed = 2)
schema(unlinked_ep2, custom_label = decode(unlinked_ep2@case_nm), seed = 2)

Labelling in diyar

Description

Encode and decode character and numeric values.

Usage

encode(x, ...)

decode(x, ...)

## Default S3 method:
encode(x, ...)

## S3 method for class 'd_label'
encode(x, ...)

## Default S3 method:
decode(x, ...)

## S3 method for class 'd_label'
decode(x, ...)

## S3 method for class 'd_label'
rep(x, ...)

## S3 method for class 'd_label'
x[i, ..., drop = TRUE]

## S3 method for class 'd_label'
x[[i, ..., drop = TRUE]]

Arguments

Details

To minimise memory usage, most components of pid, epid and pane are integer objects with labels. encode() and decode() translates these codes and labels as required.

Value

d_label; atomic

Examples

cds <- encode(rep(LETTERS[1:5], 3))
cds

nms <- decode(cds)
nms

epid object

Description

S4 objects storing the result of episodes.

Usage

is.epid(x)

as.epid(x, ...)

## S3 method for class 'epid'
format(x, ...)

## S3 method for class 'epid'
unique(x, ...)

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

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

## S3 method for class 'epid'
as.data.frame(x, ..., decode = TRUE)

## S3 method for class 'epid'
as.list(x, ..., decode = TRUE)

## S4 method for signature 'epid'
show(object)

## S4 method for signature 'epid'
rep(x, ...)

## S4 method for signature 'epid'
x[i, j, ..., drop = TRUE]

## S4 method for signature 'epid'
x[[i, j, ..., exact = TRUE]]

## S4 method for signature 'epid'
c(x, ...)

Arguments

Slots

sn

Unique record identifier.

.Data

Unique episode identifier.

wind_id

Unique reference ID for each match.

wind_nm

Type of window i.e. "Case" or "Recurrence".

case_nm

Record type in regards to case assignment.

dist_wind_index

Unit difference between each record and its window's reference record.

dist_epid_index

Unit difference between each record and its episode's reference record.

epid_dataset

Data sources in each episode.

epid_interval

The start and end dates of each episode. A number_line object.

epid_length

The duration or length of (epid_interval).

epid_total

The number of records in each episode.

iteration

The iteration when a record was matched to it's group (.Data).

options

Some options passed to the instance of episodes.

Examples

# A test for `epid` objects
ep <- episodes(date = 1)
is.epid(ep); is.epid(2)

ep <- episodes(date = 1)
is.epid(ep); is.epid(2)

Group dated events into episodes.

Description

Dated events (records) within a certain duration of an index event are assigned to a unique group. Each group has unique ID and are described as "episodes". "episodes" can be "fixed" or "rolling" ("recurring"). Each episodes has a "Case" and/or "Recurrent" record while all other records within the group are either "Duplicates" of the "Case" or "Recurrent" event.

Usage

episodes(
  date,
  case_length = Inf,
  episode_type = "fixed",
  recurrence_length = case_length,
  episode_unit = "days",
  strata = NULL,
  sn = NULL,
  episodes_max = Inf,
  rolls_max = Inf,
  case_overlap_methods = 8,
  recurrence_overlap_methods = case_overlap_methods,
  skip_if_b4_lengths = FALSE,
  data_source = NULL,
  data_links = "ANY",
  custom_sort = NULL,
  skip_order = Inf,
  reference_event = "last_record",
  case_for_recurrence = FALSE,
  from_last = FALSE,
  group_stats = c("case_nm", "wind", "epid_interval"),
  display = "none",
  case_sub_criteria = NULL,
  recurrence_sub_criteria = case_sub_criteria,
  case_length_total = 1,
  recurrence_length_total = case_length_total,
  skip_unique_strata = TRUE,
  splits_by_strata = 1,
  batched = "semi"
)

links_wf_episodes(
  date,
  case_length = Inf,
  episode_type = "fixed",
  strata = NULL,
  sn = NULL,
  display = "none"
)

episodes_af_shift(
  date,
  case_length = Inf,
  sn = NULL,
  strata = NULL,
  group_stats = FALSE,
  episode_type = "fixed",
  data_source = NULL,
  episode_unit = "days",
  data_links = "ANY",
  display = "none"
)

Arguments

Details

episodes() links dated records (events) that are within a set duration of each other in iterations. Every record is linked to a unique group (episode; epid object). These episodes represent occurrences of interest as specified by function's arguments and defined by a case definition.

Two main type of episodes are possible;

• "fixed" - An episode where all events are within a fixed duration of an index event.

• "rolling" - An episode where all events are within a recurring duration of an index event.

Every record in each episode is categorised as one of the following;

• "Case" - Index event of the episode (without a nested match criteria).

• "Case_CR" - Index event of the episode (with a nested match criteria).

• "Duplicate_C" - Duplicate of the index event.

• "Recurrent" - Recurrence of the index event (without a nested match criteria).

• "Recurrent_CR" - Recurrence of the index event (with a nested match criteria).

• "Duplicate_R" - Duplicate of the recurrent event.

• "Skipped" - Skipped records.

If data_links is supplied, every element of the list must be named "l" (links) or "g" (groups). Unnamed elements are assumed to be "l".

• If named "l", groups without records from every listed data_source will be unlinked.

• If named "g", groups without records from any listed data_source will be unlinked.

All records with a missing (NA) strata or date are skipped.

Wrapper functions or alternative implementations of episodes() for specific use cases or benefits:

• episodes_wf_splits() - Identical records are excluded from the main analysis.

• episodes_af_shift() - A mostly vectorised approach.

• links_wf_episodes() - The same functionality achieved with links.

See vignette("episodes") for further details.

Value

epid; list

See Also

episodes_wf_splits; custom_sort; sub_criteria; epid_length; epid_window; partitions; links; overlaps;

Examples

data(infections)
data(hospital_admissions)

# One 16-day (15-day difference) fixed episode per type of infection
episodes(date = infections$date,
         strata = infections$infection,
         case_length = 15,
         episodes_max = 1,
         episode_type = "fixed")

# Multiple 16-day episodes with an 11-day recurrence period
episodes(date = infections$date,
         strata = NULL,
         case_length = 15,
         episodes_max = Inf,
         episode_type = "rolling",
         recurrence_length = 10)

# Overlapping periods of hospital stays
dfr <- hospital_admissions[2:3]

dfr$admin_period <-
  number_line(dfr$admin_dt,dfr$discharge_dt)

dfr$ep <-
  episodes(date = dfr$admin_period,
           strata = NULL,
           case_length = index_window(dfr$admin_period),
           case_overlap_methods = "inbetween")

dfr
as.data.frame(dfr$ep)

Link events to chronological episodes.

Description

episodes_wf_splits is a wrapper function of episodes. It's designed to be more efficient with larger datasets. Duplicate records which do not affect the case definition are excluded prior to episode tracking. The resulting episode identifiers are then recycled for the duplicate records.

Usage

episodes_wf_splits(..., duplicates_recovered = "ANY", reframe = FALSE)

Arguments

Details

episodes_wf_splits() reduces or re-frames a dataset to the minimum datasets required to implement a case definition. This leads to the same outcome but with the benefit of a shorter processing time.

The duplicates_recovered argument determines which identifiers are recycled. Selecting the "with_sub_criteria" option will force only identifiers created resulting from a matched sub_criteria ("Case_CR" and "Recurrent_CR") are recycled. However, if "without_sub_criteria" is selected then only identifiers created that do not result from a matched sub_criteria ("Case" and "Recurrent") are recycled Excluded duplicates of "Duplicate_C" and "Duplicate_R" are always recycled.

The reframe argument will either reframe or subset a sub_criteria. Both will require slightly different functions for match_funcs or equal_funcs.

Value

epid; list

See Also

episodes; sub_criteria

Examples

# With 2,000 duplicate records of 20 events,
# `episodes_wf_splits()` will take less time than `episodes()`
dates <- seq(from = as.Date("2019-04-01"), to = as.Date("2019-04-20"), by = 1)
dates <- rep(dates, 2000)

system.time(
  ep1 <- episodes(dates, 1)
)
system.time(
  ep2 <- episodes_wf_splits(dates, 1)
)

# Both leads to the same outcome.
all(ep1 == ep2)

Record linkage

Description

Deterministic and probabilistic record linkage Assign unique identifiers to records based on partial, nested or calculated probabilities.

Usage

links_af_probabilistic(
  attribute,
  blocking_attribute = NULL,
  cmp_func = diyar::exact_match,
  attr_threshold = 1,
  probabilistic = TRUE,
  m_probability = 0.95,
  u_probability = NULL,
  score_threshold = 1,
  repeats_allowed = FALSE,
  permutations_allowed = FALSE,
  data_source = NULL,
  ignore_same_source = TRUE,
  display = "none"
)

links_wf_probabilistic(
  attribute,
  blocking_attribute = NULL,
  cmp_func = diyar::exact_match,
  attr_threshold = 1,
  probabilistic = TRUE,
  m_probability = 0.95,
  u_probability = NULL,
  score_threshold = 1,
  id_1 = NULL,
  id_2 = NULL,
  return_weights = FALSE,
  ...
)

prob_score_range(attribute, m_probability = 0.95, u_probability = NULL)

Arguments

Details

links_wf_probabilistic() - A wrapper function of links with a with a specific sub_criteria and to achieve to achieve probabilistic record linkage It excludes functionalities for the nested and multi-stage linkage. links_wf_probabilistic() requires a score_threshold in advance. To help with this, prob_score_range() can be used to return the range of scores attainable for a given set of attribute, m and u-probabilities. Additionally, id_1 and id_2 can be used to link specific records pairs, aiding the review of potential scores.

links_af_probabilistic() - A simpler version of links. It excludes functionalities for the batched, nested and multi-stage linkage. links_af_probabilistic() requires a score_threshold in advance, however, since it exports the match weights, the score_threshold can be changed after the analysis.

Value

pid; list

References

Fellegi, I. P., & Sunter, A. B. (1969). A Theory for Record Linkage. Journal of the Statistical Association, 64(328), 1183 - 1210. https://doi.org/10.1080/01621459.1969.10501049

Asher, J., Resnick, D., Brite, J., Brackbill, R., & Cone, J. (2020). An Introduction to Probabilistic Record Linkage with a Focus on Linkage Processing for WTC Registries. International journal of environmental research and public health, 17(18), 6937. https://doi.org/10.3390/ijerph17186937. See vignette("links") for more information.

See Also

links

Examples

data(patient_records)
# Weighted (probabilistic) comparison of forename, middlename and surname
criteria_1 <- as.list(patient_records[c("forename", "middlename", "surname")])

# Possible scores when m-probability is 0.95
prob_scores <- prob_score_range(attribute = criteria_1,
                                m_probability = 0.95,
                                u_probability = NULL)
## Not run: 
# Probabilistic record linkage with 'links_af_probabilistic()'
pids_1a <- links_af_probabilistic(attribute = criteria_1,
                        cmp_func = exact_match,
                        attr_threshold = 1,
                        probabilistic = TRUE,
                        m_probability = 0.95,
                        score_threshold = prob_scores$mid_scorce,
                        display = "stats")

# Equivalent with 'links_wf_probabilistic()'
pids_1b <- links_wf_probabilistic(attribute = criteria_1,
                                  cmp_func = exact_match,
                                  attr_threshold = 1,
                                  probabilistic = TRUE,
                                  m_probability = 0.95,
                                  score_threshold = prob_scores$mid_scorce,
                                  display = "progress",
                                  recursive = TRUE,
                                  check_duplicates = TRUE)

# Less thorough but faster equivalent with `links_wf_probabilistic()`
pids_1c <- links_wf_probabilistic(attribute = criteria_1,
                                  cmp_func = exact_match,
                                  attr_threshold = 1,
                                  probabilistic = TRUE,
                                  m_probability = 0.95,
                                  score_threshold = prob_scores$mid_scorce,
                                  display = "progress",
                                  recursive = FALSE,
                                  check_duplicates = FALSE)

# Each implementation can lead to different results
summary(pids_1a$pid)
summary(pids_1b$pid)
summary(pids_1c$pid)

## End(Not run)

# Weighted (non-probabilistic) comparison of forename, middlename and age difference
criteria_2 <- as.list(patient_records[c("forename", "middlename", "dateofbirth")])
age_diff <- function(x, y){
  diff <- abs(as.numeric(x) - as.numeric(y))
  wgt <-  diff %in% 0:(365 * 10) & !is.na(diff)
  wgt
}

pids_2a <- links_af_probabilistic(attribute = criteria_2,
                        blocking_attribute = patient_records$surname,
                        cmp_func = c(exact_match, exact_match, age_diff),
                        score_threshold = number_line(3, 5),
                        probabilistic = FALSE,
                        display = "stats")

# Larger weights can be assigned to particular attributes through `cmp_func`
# For example, a smaller age difference can contribute a higher score (e.g 0 to 3)
age_diff_2 <- function(x, y){
  diff <- as.numeric(abs(x - y))
  wgt <-  diff %in% 0:(365 * 10) & !is.na(diff)
  wgt[wgt] <- match(as.numeric(cut(diff[wgt], 3)), 3:1)
  wgt
}
pids_2b <- links_af_probabilistic(attribute = criteria_2,
                        blocking_attribute = patient_records$surname,
                        cmp_func = c(exact_match, exact_match, age_diff_2),
                        score_threshold = number_line(3, 5),
                        probabilistic = FALSE,
                        display = "stats")

head(pids_2a$pid_weights, 10)
head(pids_2b$pid_weights, 10)

Multistage record linkage

Description

Assign records to unique groups based on an ordered set of match criteria.

Usage

links(
  criteria,
  sub_criteria = NULL,
  sn = NULL,
  strata = NULL,
  data_source = NULL,
  data_links = "ANY",
  display = "none",
  group_stats = FALSE,
  expand = TRUE,
  shrink = FALSE,
  recursive = "none",
  check_duplicates = FALSE,
  tie_sort = NULL,
  batched = "yes",
  repeats_allowed = FALSE,
  permutations_allowed = FALSE,
  ignore_same_source = FALSE
)

Arguments

Details

The priority of matches decreases with each subsequent stage of the linkage process. Therefore, the attributes in criteria should be in an order of decreasing relevance.

Records with missing data (NA) for each criteria are skipped at the respective stage, while records with missing data strata are skipped from every stage.

If a record is skipped from a stage, another attempt will be made to match the record at the next stage. If a record is still unmatched by the last stage, it is assigned a unique group ID.

A sub_criteria adds nested match criteria to each stage of the linkage process. If used, only records with a matching criteria and sub_criteria are linked.

In links, each sub_criteria must be linked to a criteria. This is done by adding each sub_criteria to a named element of a list - "cr" concatenated with the corresponding stage's number. For example, 3 sub_criteria linked to criteria 1, 5 and 13 will be;

list(cr1 = sub_criteria(...), cr5 = sub_criteria(...), cr13 = sub_criteria(...))

Any unlinked sub_criteria will be ignored.

Every element in data_links must be named "l" (links) or "g" (groups). Unnamed elements of data_links will be assumed to be "l".

• If named "l", groups without records from every listed data_source will be unlinked.

• If named "g", groups without records from any listed data_source will be unlinked.

See vignette("links") for more information.

Value

pid; list

See Also

links_af_probabilistic; episodes; predefined_tests; sub_criteria

Examples

data(patient_records)
dfr <- patient_records
# An exact match on surname followed by an exact match on forename
stages <- as.list(dfr[c("surname", "forename")])
p1 <- links(criteria = stages)

# An exact match on forename followed by an exact match on surname
p2 <- links(criteria = rev(stages))

# Nested matches
# Same sex OR birth year
m.cri.1 <- sub_criteria(
  format(dfr$dateofbirth, "%Y"), dfr$sex,
  operator = "or")

# Same middle name AND a 10 year age difference
age_diff <- function(x, y){
  diff <- abs(as.numeric(x) - as.numeric(y))
  wgt <-  diff %in% 0:10 & !is.na(diff)
  wgt
}
m.cri.2 <- sub_criteria(
  format(dfr$dateofbirth, "%Y"), dfr$middlename,
  operator = "and",
  match_funcs = c(age_diff, exact_match))

# Nested match criteria 'm.cri.1' OR 'm.cri.2'
n.cri <- sub_criteria(
  m.cri.1, m.cri.2,
  operator = "or")

# Record linkage with additional match criteria
p3 <- links(
  criteria = stages,
  sub_criteria = list(cr1 = m.cri.1,
                      cr2 = m.cri.2))

# Record linkage with additonal nested match criteria
p4 <- links(
  criteria = stages,
  sub_criteria = list(cr1 = n.cri,
                      cr2 = n.cri))

dfr$p1 <- p1; dfr$p2 <- p2
dfr$p3 <- p3; dfr$p4 <- p4

head(dfr)

Grammatical lists.

Description

A convenience function to format atomic vectors as a written list.

Usage

listr(x, sep = ", ", conj = " and ", lim = Inf)

Arguments

Value

character.

Examples

listr(1:5)
listr(1:5, sep = "; ")
listr(1:5, sep = "; ", conj = " and")
listr(1:5, sep = "; ", conj = " and", lim = 2)

Convert an edge list to record identifiers.

Description

Convert an edge list to record identifiers.

Usage

make_ids(x_pos, y_pos, id_length = max(x_pos, y_pos))

Arguments

Details

Record groups from non-recursive links have the lowest record ID (sn) in the set as their group ID.

Value

list

Examples

make_ids(x_pos = rep(7, 7), y_pos = 1:7)
make_ids(x_pos = c(1, 6), y_pos = 6:7)
make_ids(x_pos = 1:5, y_pos = c(1, 1, 2, 3, 4))

Combinations and permutations of record-sets.

Description

Combinations and permutations of record-sets.

Usage

sets(n, r, permutations_allowed = TRUE, repeats_allowed = TRUE)

make_sets(
  x,
  r,
  strata = NULL,
  permutations_allowed = TRUE,
  repeats_allowed = TRUE
)

make_pairs(
  x,
  strata = NULL,
  repeats_allowed = TRUE,
  permutations_allowed = FALSE
)

make_pairs_wf_source(..., data_source = NULL)

Arguments

Details

set() - Create r-set combinations or permutations of n observations.

make_set() - Create r-set combinations or permutations of vector x.

make_pairs() - Create 2-set combinations or permutations of vector x.

make_pairs_wf_source() - Create 2-set combinations or permutations of vector x that are from different sources (data_source).

Value

A list of a vector's elements and corresponding indexes.

See Also

eval_sub_criteria

Examples

sets(4, 2)
sets(4, 2, repeats_allowed = FALSE, permutations_allowed = FALSE)
make_sets(month.abb[1:4], 2)
make_sets(month.abb[1:4], 3)

make_pairs(month.abb[1:4])
make_pairs(month.abb[1:4], strata = c(1, 1, 2, 2))
make_pairs_wf_source(month.abb[1:4], data_source = c(1, 1, 2, 2))

Create epid and pid objects with index of matching records

Description

Create epid and pid objects with index of matching records

Usage

make_episodes(
  x_pos,
  y_pos,
  x_val,
  date,
  case_nm,
  wind_id,
  wind_nm,
  from_last,
  data_source,
  data_links,
  iteration,
  options,
  episode_unit
)

make_pids(
  x_pos,
  y_pos,
  x_val,
  link_id,
  pid_cri,
  data_source,
  data_links,
  iteration
)

Arguments

Merge group identifiers

Description

Consolidate two group identifiers.

Usage

merge_ids(...)

## Default S3 method:
merge_ids(id1, id2, tie_sort = NULL, expand = TRUE, shrink = FALSE, ...)

## S3 method for class 'pid'
merge_ids(id1, id2, tie_sort = NULL, expand = TRUE, shrink = FALSE, ...)

## S3 method for class 'epid'
merge_ids(id1, id2, tie_sort = NULL, expand = TRUE, shrink = FALSE, ...)

## S3 method for class 'pane'
merge_ids(id1, id2, tie_sort = NULL, expand = TRUE, shrink = FALSE, ...)

Arguments

Details

Groups in id1 are expanded or shrunk by groups in id2.

A unique group with only one record is considered a non-matching record.

Note that the expand and shrink features are not interchangeable. The outcome when shrink is TRUE is not the same when expand is FALSE. See Examples.

See Also

links; links_af_probabilistic

Examples

id1 <- rep(1, 5)
id2 <- c(2, 2, 3, 3, 3)
merge_ids(id1, id2, shrink = TRUE)

id1 <- c(rep(1, 3), 6, 7)
id2 <- c(2,2,3,3,3)
merge_ids(id1, id2, shrink = TRUE)
merge_ids(id1, id2, expand = FALSE)

id1 <- rep(1, 5)
id2 <- c(1:3, 4, 4)
merge_ids(id1, id2, shrink = TRUE)
merge_ids(id1, id2, expand= FALSE)

data(missing_staff_id)
dfr <- missing_staff_id
id1 <- links(dfr[[5]])
id2 <- links(dfr[[6]])
merge_ids(id1, id2)

number_line

Description

A range of numeric values.

Usage

number_line(l, r, id = NULL, gid = NULL)

as.number_line(x)

is.number_line(x)

left_point(x)

left_point(x) <- value

right_point(x)

right_point(x) <- value

start_point(x)

start_point(x) <- value

end_point(x)

end_point(x) <- value

number_line_width(x)

reverse_number_line(x, direction = "both")

shift_number_line(x, by = 1)

expand_number_line(x, by = 1, point = "both")

invert_number_line(x, point = "both")

number_line_sequence(
  x,
  by = NULL,
  length.out = 1,
  fill = TRUE,
  simplify = FALSE
)

Arguments

Details

A number_line object represents a range of numbers. It is made up of a start and end point as the lower and upper ends of the range respectively. The location of the start point - left or right, determines whether it is an "increasing" or "decreasing" number_line. This is the direction of the number_line.

reverse_number_line() - reverse the direction of a number_line. A reversed number_line has its left and right points swapped. The direction argument specifies which type of number_line will be reversed. number_line with non-finite start or end points (i.e. NA, NaN and Inf) can't be reversed.

shift_number_line() - Shift a number_line towards the positive or negative end of the number line.

expand_number_line() - Increase or decrease the width of a number_line.

invert_number_line() - Change the left or right points from a negative to positive value or vice versa.

number_line_sequence() - Split a number_line into equal parts (length.out) or by a fixed recurring width (by).

Value

number_line

See Also

overlaps; set_operations; episodes; links

Examples

number_line(-100, 100)

# Also compatible with other numeric based object classes
number_line(as.POSIXct("2019-05-15 13:15:07", tz = "UTC"),
            as.POSIXct("2019-05-15 15:17:10", tz = "UTC"))

# Coerce compatible object classes to `number_line` objects
as.number_line(5.1); as.number_line(as.Date("2019-10-21"))

# A test for number_line objects
a <- number_line(as.Date("2019-04-25"), as.Date("2019-01-01"))
is.number_line(a)

# Structure of a number_line object
left_point(a); right_point(a); start_point(a); end_point(a)

# Reverse number_line objects
reverse_number_line(number_line(as.Date("2019-04-25"), as.Date("2019-01-01")))
reverse_number_line(number_line(200, -100), "increasing")
reverse_number_line(number_line(200, -100), "decreasing")

c <- number_line(5, 6)
# Shift number_line objects towards the positive end of the number line
shift_number_line(x = c(c, c), by = c(2, 3))
# Shift number_line objects towards the negative end of the number line
shift_number_line(x = c(c, c), by = c(-2, -3))

# Change the duration, width or length of a number_line object
d <- c(number_line(3, 6), number_line(6, 3))

expand_number_line(d, 2)
expand_number_line(d, -2)
expand_number_line(d, c(2,-1))
expand_number_line(d, 2, "start")
expand_number_line(d, 2, "end")

# Invert `number_line` objects
e <- c(number_line(3, 6), number_line(-3, -6), number_line(-3, 6))
e
invert_number_line(e)
invert_number_line(e, "start")
invert_number_line(e, "end")

# Split number line objects
x <- number_line(Sys.Date() - 5, Sys.Date())
x
number_line_sequence(x, by = 2)
number_line_sequence(x, by = 4)
number_line_sequence(x, by = 4, fill = FALSE)
number_line_sequence(x, length.out = 2)

number_line object

Description

S4 objects representing a range of numeric values

Usage

## S4 method for signature 'number_line'
show(object)

## S4 method for signature 'number_line'
rep(x, ...)

## S4 method for signature 'number_line'
x[i, j, ..., drop = TRUE]

## S4 method for signature 'number_line'
x[[i, j, ..., exact = TRUE]]

## S4 replacement method for signature 'number_line'
x[i, j, ...] <- value

## S4 replacement method for signature 'number_line'
x[[i, j, ...]] <- value

## S4 method for signature 'number_line'
x$name

## S4 replacement method for signature 'number_line'
x$name <- value

## S4 method for signature 'number_line'
c(x, ...)

## S3 method for class 'number_line'
unique(x, ...)

## S3 method for class 'number_line'
seq(x, precision = NULL, fill = FALSE, ...)

## S3 method for class 'number_line'
sort(x, decreasing = FALSE, ...)

## S3 method for class 'number_line'
format(x, ...)

## S3 method for class 'number_line'
as.list(x, ...)

## S3 method for class 'number_line'
as.data.frame(x, ...)

Arguments

Slots

start

First value in the range.

id

Unique element id. Optional.

gid

Unique group id. Optional.

.Data

Length, duration or width of the range.

Overlapping number line objects

Description

Identify overlapping number_line objects

Usage

overlaps(x, y, methods = 8)

overlap(x, y)

none(x, y)

exact(x, y)

across(x, y)

x_across_y(x, y)

y_across_x(x, y)

chain(x, y)

x_chain_y(x, y)

y_chain_x(x, y)

aligns_start(x, y)

x_aligns_start_y(x, y)

y_aligns_start_x(x, y)

aligns_end(x, y)

x_aligns_end_y(x, y)

y_aligns_end_x(x, y)

inbetween(x, y)

x_inbetween_y(x, y)

y_inbetween_x(x, y)

overlap_method(x, y)

include_overlap_method(methods)

exclude_overlap_method(methods)

overlap_method_codes(methods)

overlap_method_names(methods)

Arguments

Details

There are 6 mutually exclusive types of overlap;

• exact() - identical start_point and end_point points.

• inbetween() - Both start_point and end_point of one number_line object are within the start_point and end_point of another.

• across() - Only the start_point or end_point of one number_line object is in between the start_point and end_point of another.

• chain() - end_point of one number_line object is identical to the start_point of another.

• aligns_start() - identical start_point only.

• aligns_end() - identical end_point only.

Except exact(), each type of overlap has two variations;

• x_`method`_y() - number_line-x starts before number_line-y.

• y_`method`_x() - number_line-y starts before number_line-x.

There are two mutually inclusive types of overlap;

• overlap() - a convenient option to select "ANY" and "ALL" type of overlap.

• none() - a convenient option to select "NO" type of overlap.

Selecting multiple types of overlap;

• overlaps() - select specific type(s) of overlap.

• overlap_method() - return the type of overlap for a pair of number_line objects.

• overlap_method_codes() - return the corresponding overlap method code for a specific type(s) of overlap.

• overlap_method_names() - return the corresponding type(s) of overlap for a specific overlap code.

• include_overlap_method() - return a character(1) value for specified type(s) of overlap.

• exclude_overlap_method() - return a character(1) value for all type(s) of overlap except those specified.

Value

logical; character

See Also

number_line; set_operations

Examples

a <- number_line(-100, 100)
g <- number_line(100, 100)
overlaps(a, g)

# It's neither an "exact" or "chain"-overlap
overlaps(a, g, methods = "exact|chain")

# It's an "aligns_end"-overlap
overlap_method(a, g)
overlaps(a, g, methods = "exact|chain|x_aligns_end_y")

# Corresponding overlap code
overlap_method_codes("exact|chain|x_aligns_end_y")
include_overlap_method(c("exact", "chain", "x_aligns_end_y"))

# Corresponding overlap name
overlap_method_names(overlap_method_codes("exact|chain|x_aligns_end_y"))

# Every other type overlap
exclude_overlap_method(c("exact", "chain", "x_aligns_end_y"))
overlap_method_names(exclude_overlap_method(c("exact", "chain", "x_aligns_end_y")))

# All the above is based on tests for each specific type of overlap as seen below
none(a, g)
exact(a, g)
across(a, g)
x_across_y(a, g)
y_across_x(a, g)
chain(a, g)
x_chain_y(a, g)
y_chain_x(a, g)
inbetween(a, g)
x_inbetween_y(a, g)
y_inbetween_x(a, g)
aligns_start(a, g)
x_aligns_start_y(a, g)
y_aligns_start_x(a, g)
aligns_end(a, g)
x_aligns_end_y(a, g)
y_aligns_end_x(a, g)

pane object

Description

S4 objects storing the result of partitions.

Usage

is.pane(x)

as.pane(x)

## S3 method for class 'pane'
format(x, ...)

## S3 method for class 'pane'
unique(x, ...)

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

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

## S3 method for class 'pane'
as.data.frame(x, ..., decode = TRUE)

## S3 method for class 'pane'
as.list(x, ..., decode = TRUE)

## S4 method for signature 'pane'
show(object)

## S4 method for signature 'pane'
rep(x, ...)

## S4 method for signature 'pane'
x[i, j, ..., drop = TRUE]

## S4 method for signature 'pane'
x[[i, j, ..., exact = TRUE]]

## S4 method for signature 'pane'
c(x, ...)

Arguments

Slots

sn

Unique record identifier.

.Data

Unique pane identifier.

case_nm

Record type in regards to index assignment.

window_list

A list of considered windows for each pane.

dist_pane_index

The difference between each event and it's index event.

pane_dataset

Data sources in each pane.

pane_interval

The start and end dates of each pane. A number_line object.

pane_length

The duration or length of (pane_interval).

pane_total

The number of records in each pane.

options

Some options passed to the instance of partitions.

window_matched

A list of matched windows for each pane.

Examples

# A test for pane objects
pn <- partitions(date = 1, by = 1)
is.pane(pn); is.pane(2)

Distribute events into specified intervals.

Description

Distribute events into groups defined by time or numerical intervals. Each set of linked records are assigned a unique identifier with relevant group-level data.

Usage

partitions(
  date,
  window = NULL,
  windows_total = 1,
  separate = FALSE,
  sn = NULL,
  strata = NULL,
  data_links = "ANY",
  custom_sort = NULL,
  group_stats = FALSE,
  data_source = NULL,
  by = NULL,
  length.out = NULL,
  fill = TRUE,
  display = "none",
  precision = 1
)

Arguments

Details

Each assigned group is referred to as a pane A pane consists of events within a specific time or numerical intervals (window).

Each window must cover a separate interval. Overlapping windows are merged before events are distributed into panes. Events that occur over two windows are assigned to the last one listed.

Alternatively, you can create windows by splitting a period into equal parts (length.out), or into a sequence of intervals with fixed widths (by).

By default, the earliest event is taken as the "Index" event of the pane. An alternative can be chosen with custom_sort. Note that this is simply a convenience option because it has no bearing on how groups are assigned.

partitions() will categorise records into 3 types;

• "Index" - Index event/record of the pane.

• "Duplicate_I" - Duplicate of the "Index" record.

• "Skipped" - Records that are not assigned to a pane.

Every element in data_links must be named "l" (links) or "g" (groups). Unnamed elements of data_links will be assumed to be "l".

• If named "l", only groups with records from every listed data_source will be retained.

• If named "g", only groups with records from any listed data_source will be retained.

NA values in strata excludes records from the partitioning process.

See vignette("episodes") for more information.

Value

pane

See Also

pane; number_line_sequence; episodes; links; overlaps; number_line; schema

Examples

events <- c(30, 2, 11, 10, 100)
windows <- number_line(c(1, 9, 25), c(3, 12, 35))

events
partitions(date = events, length.out = 3, separate = TRUE)
partitions(date = events, by = 10, separate = TRUE)
partitions(date = events, window = windows, separate = TRUE)
partitions(date = events, window = windows, separate = FALSE)
partitions(date = events, window = windows, separate = FALSE, windows_total = 4)

pid objects

Description

S4 objects storing the result of links.

Usage

is.pid(x)

as.pid(x, ...)

## S3 method for class 'pid'
format(x, ...)

## S3 method for class 'pid'
unique(x, ...)

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

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

## S3 method for class 'pid'
as.data.frame(x, ..., decode = TRUE)

## S3 method for class 'pid'
as.list(x, ..., decode = TRUE)

## S4 method for signature 'pid'
show(object)

## S4 method for signature 'pid'
rep(x, ...)

## S4 method for signature 'pid'
x[i, j, ..., drop = TRUE]

## S4 method for signature 'pid'
x[[i, j, ..., exact = TRUE]]

## S4 method for signature 'pid'
c(x, ...)

Arguments

Slots

sn

Unique record identifier.

.Data

Unique group identifier.

link_id

Unique reference ID for each match.

pid_cri

Match stage of the step-wise linkage.

pid_dataset

Data sources in each group.

pid_total

The number of records in each group.

iteration

The iteration when a record was matched to it's group (.Data).

Examples

# A test for pid objects
pd <- links(criteria = 1)
is.pid(pd); is.pid(2)

Predefined logical tests in diyar

Description

A collection of predefined logical tests used with sub_criteria objects

Usage

exact_match(x, y)

range_match(x, y, range = 10)

prob_link(
  x,
  y,
  cmp_func,
  attr_threshold,
  score_threshold,
  probabilistic,
  return_weights = FALSE
)

true(x, y)

false(x, y)

Arguments

Details

exact_match() - test that x == y

range_match() - test that x \le y \le (x + range)

prob_link() - Test that a record-pair relate to the same entity based on Fellegi and Sunter (1969) model for deciding if two records belong to the same entity.

In summary, record-pairs are created and categorised as matches and non-matches (attr_threshold) with user-defined functions (cmp_func). If probabilistic is TRUE, two probabilities (m and u) are used to calculate weights for matches and non-matches. The m-probability is the probability that matched records are actually from the same entity i.e. a true match, while u-probability is the probability that matched records are not from the same entity i.e. a false match. Record-pairs whose total score are above a certain threshold (score_threshold) are assumed to belong to the same entity.

Agreement (match) and disagreement (non-match) scores are calculated as described by Asher et al. (2020).

For each record pair, an agreement for attribute i is calculated as;

\log_{2}(m_{i}/u_{i})

For each record pair, a disagreement score for attribute i is calculated as;

\log_{2}((1-m_{i})/(1-u_{i}))

where m_{i} and u_{i} are the m and u-probabilities for each value of attribute i.

Note that each probability is calculated as a combined probability for the record pair. For example, if the values of the record-pair have u-probabilities of 0.1 and 0.2 respectively, then the u-probability for the pair will be 0.02.

Missing data (NA) are considered non-matches and assigned a u-probability of 0.

Examples

`exact_match`
exact_match(x = 1, y = 1)
exact_match(x = 1, y = 2)

`range_match`
range_match(x = 10, y = 16, range = 6)
range_match(x = 16, y = 10, range = 6)

Modify sub_criteria objects

Description

Modify the attributes of a sub_criteria object.

Usage

reframe(x, ...)

## S3 method for class 'sub_criteria'
reframe(x, func = identity, ...)

Arguments

See Also

sub_criteria; eval_sub_criteria; attr_eval

Examples

s_cri <- sub_criteria(month.abb, month.name)
reframe(s_cri, func = function(x) x[12])
reframe(s_cri, func = function(x) x[12:1])
reframe(s_cri, func = function(x) attrs(x[1:6], x[7:12]))

Schema diagram for group identifiers

Description

Create schema diagrams for number_line, epid, pid and pane objects.

Usage

schema(x, ...)

## S3 method for class 'number_line'
schema(x, show_labels = c("date", "case_overlap_methods"), ...)

## S3 method for class 'epid'
schema(
  x,
  title = NULL,
  show_labels = c("length_arrow"),
  show_skipped = TRUE,
  show_non_finite = FALSE,
  theme = "dark",
  seed = NULL,
  custom_label = NULL,
  ...
)

## S3 method for class 'pane'
schema(
  x,
  title = NULL,
  show_labels = c("window_label"),
  theme = "dark",
  seed = NULL,
  custom_label = NULL,
  ...
)

## S3 method for class 'pid'
schema(
  x,
  title = NULL,
  show_labels = TRUE,
  theme = "dark",
  orientation = "by_pid",
  seed = NULL,
  custom_label = NULL,
  ...
)

Arguments

Details

A visual aid to describe the data linkage (links), episode tracking (episodes) or partitioning process (partitions).

show_labels options (multi-select)

• schema.epid - TRUE, FALSE, "sn", "epid", "date", "case_nm", "wind_nm", "length", "length_arrow", "case_overlap_methods" or "recurrence_overlap_methods"

• schema.pane - TRUE, FALSE, "sn", "pane", "date", "case_nm" or "window_label"

• schema.pid - TRUE, FALSE, "sn" or "pid"

Value

ggplot objects

Examples

schema(number_line(c(1, 2), c(2, 1)))

schema(episodes(1:10, 2))

schema(partitions(1:10, by = 2, separate = TRUE))

schema(links(list(c(1, 1, NA, NA), c(NA, 1, 1, NA))))

Set operations on number line objects

Description

Perform set operations on a pair of [number_line]s.

Usage

union_number_lines(x, y)

intersect_number_lines(x, y)

subtract_number_lines(x, y)

Arguments

Details

union_number_lines() - Combined the range of x and that of y

intersect_number_line() - Subset of x that overlaps with y and vice versa

subtract_number_lines() - Subset of x that does not overlap with y and vice versa.

The direction of the returned [number_line] will be that of the widest one (x or y). If x and y have the same length, it'll be an "increasing" direction.

If x and y do not overlap, NA ("NA ?? NA") is returned.

Value

[number_line]; list

See Also

number_line; overlaps

Examples

nl_1 <- c(number_line(1, 5), number_line(1, 5), number_line(5, 9))
nl_2 <- c(number_line(1, 2), number_line(2, 3), number_line(0, 6))

# Union
nl_1; nl_2; union_number_lines(nl_1, nl_2)


nl_3 <- number_line(as.Date(c("01/01/2020", "03/01/2020","09/01/2020"), "%d/%m/%Y"),
                    as.Date(c("09/01/2020", "09/01/2020","25/12/2020"), "%d/%m/%Y"))

nl_4 <- number_line(as.Date(c("04/01/2020","01/01/2020","01/01/2020"), "%d/%m/%Y"),
                    as.Date(c("05/01/2020","05/01/2020","03/01/2020"), "%d/%m/%Y"))

# Intersect
nl_3; nl_4; intersect_number_lines(nl_3, nl_4)

# Subtract
nl_3; nl_4; subtract_number_lines(nl_3, nl_4)

Datasets in diyar package

Description

Datasets in diyar package

Usage

data(staff_records)

data(missing_staff_id)

data(infections)

data(infections_2)

data(infections_3)

data(infections_4)

data(hospital_admissions)

data(patient_list)

data(patient_list_2)

data(hourly_data)

data(Opes)

data(episode_unit)

data(overlap_methods)

data(patient_records)

Format

data.frame

data.frame

data.frame

data.frame

data.frame

data.frame

data.frame

data.frame

An object of class data.frame with 5 rows and 4 columns.

data.frame

data.frame

list

list

data.frame

Details

staff_records - Staff record with some missing data

missing_staff_id - Staff records with missing staff identifiers

infections, infections_2, infections_3 and infections_4 - Reports of bacterial infections

hospital_admissions - Hospital admissions and discharges

patient_list & patient_list_2 - Patient list with some missing data

Hourly data

Opes - List of individuals with the same name

Duration in seconds for each 'episode_unit'

Permutations of number_line overlap methods

Examples

data(staff_records)
data(missing_staff_id)
data(infections)
data(infections_2)
data(infections_3)
data(infections_4)
data(hospital_admissions)
data(patient_list)
data(patient_list_2)
data(hourly_data)
data(Opes)
data(episode_unit)
data(overlap_methods)
data(patient_records)

Match criteria

Description

Match criteria for record linkage with links and episodes

Usage

sub_criteria(
  ...,
  match_funcs = c(exact = diyar::exact_match),
  equal_funcs = c(exact = diyar::exact_match),
  operator = "or"
)

attrs(..., .obj = NULL)

eval_sub_criteria(x, ...)

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

## S3 method for class 'sub_criteria'
format(x, show_levels = FALSE, ...)

## S3 method for class 'sub_criteria'
eval_sub_criteria(
  x,
  x_pos = seq_len(max(attr_eval(x))),
  y_pos = rep(1L, length(x_pos)),
  check_duplicates = TRUE,
  depth = 0,
  ...
)

Arguments

Details

sub_criteria() - Create a match criteria as a sub_criteria object. A sub_criteria object contains attributes to be compared, logical tests for the comparisons (see predefined_tests for examples) and another set of logical tests to determine identical records.

attrs() - Create a d_attribute object - a collection of atomic objects that can be passed to sub_criteria() as a single attribute.

eval_sub_criteria() - Evaluates a sub_criteria object.

At each iteration of links or episodes, record-pairs are created from each attribute of a sub_criteria object. eval_sub_criteria() evaluates each record-pair using the match_funcs and equal_funcs functions of a sub_criteria object. See predefined_tests for examples of match_funcs and equal_funcs.

User-defined functions are also permitted as match_funcs and equal_funcs. Such functions must meet three requirements:

1. It must be able to compare the attributes.

2. It must have two arguments named `x` and `y`, where `y` is the value for one observation being compared against all other observations (`x`).

3. It must return a logical object i.e. TRUE or FALSE.

attrs() is useful when the match criteria requires an interaction between the multiple attributes. For example, attribute 1 + attribute 2 > attribute 3.

Every attribute, including those in attrs(), must have the same length or a length of 1.

Value

sub_criteria

See Also

predefined_tests; links; episodes; eval_sub_criteria

Examples

# Attributes
attr_1 <- c(30, 28, 40, 25, 25, 29, 27)
attr_2 <- c("M", "F", "U", "M", "F", "U", "M")

# A match criteria
## Example 1 - A maximum difference of 10 in attribute 1
s_cri1 <- sub_criteria(attr_1, match_funcs = range_match)
s_cri1

# Evaluate the match criteria
## Compare the first element of 'attr_1' against all other elements
eval_sub_criteria(s_cri1)
## Compare the second element of 'attr_1' against all other elements
x_pos_val <- seq_len(max(attr_eval(s_cri1)))
eval_sub_criteria(s_cri1,
                  x_pos = x_pos_val,
                  y_pos = rep(2, length(x_pos_val)))

## Example 2 - `s_cri1` AND an exact match on attribute 2
s_cri2 <- sub_criteria(
  s_cri1,
  sub_criteria(attr_2, match_funcs = exact_match),
  operator = "and")
s_cri2

## Example 3 - `s_cri1` OR an exact match on attribute 2
s_cri3 <- sub_criteria(
  s_cri1,
  sub_criteria(attr_2, match_funcs = exact_match),
  operator = "or")
s_cri3

# Evaluate the match criteria
eval_sub_criteria(s_cri2)
eval_sub_criteria(s_cri3)

# Alternatively, using `attr()`
AND_func <- function(x, y) range_match(x$a1, y$a1) & x$a2 == y$a2
OR_func <- function(x, y) range_match(x$a1, y$a1) | x$a2 == y$a2

## Create a match criteria
s_cri2b <- sub_criteria(attrs(.obj = list(a1 = attr_1, a2 = attr_2)),
                        match_funcs = AND_func)
s_cri3b <- sub_criteria(attrs(.obj = list(a1 = attr_1, a2 = attr_2)),
                        match_funcs = OR_func)

# Evaluate the match criteria
eval_sub_criteria(s_cri2b)
eval_sub_criteria(s_cri3b)

Windows and lengths

Description

Covert windows to and from case_lengths and recurrence_lengths.

Usage

epid_windows(date, lengths, episode_unit = "days")

epid_lengths(date, windows, episode_unit = "days")

index_window(date, from_last = FALSE)

Arguments

Details

epid_windows - returns the corresponding window for a given a date, and case_length or recurrence_length.

epid_lengths - returns the corresponding case_length or recurrence_length for a given date and window.

index_window - returns the corresponding case_length or recurrence_length for the date only.

index_window(date = x) is a convenience function for epid_lengths(date = x, window = x).

Value

number_line.

Examples

# Which `window` will a given `length` cover?
date <- Sys.Date()
epid_windows(date, 10)
epid_windows(date, number_line(5, 10))
epid_windows(date, number_line(-5, 10))
epid_windows(date, -5)


# Which `length` is required to cover a given `window`?
date <- number_line(Sys.Date(), Sys.Date() + 20)
epid_lengths(date, Sys.Date() + 30)
epid_lengths(date, number_line(Sys.Date() + 25, Sys.Date() + 30))
epid_lengths(date, number_line(Sys.Date() - 10, Sys.Date() + 30))
epid_lengths(date, Sys.Date() - 10)

# Which `length` is required to cover the `date`?
index_window(20)
index_window(number_line(15, 20))