pattern_search: pattern_search

In ilangurudev/approxmapR: ApproxMAP (APPROXimate Multiple Alignment Pattern)

pattern_search

Description

A function which finds patterns and has the ability to handle regular expression using the R syntax. For ease, some shorthands have been created for common pattern structures.

Usage

  pattern_search <- function(Clustered_Dataframe, find_pattern = NULL, event_set = FALSE, exact = FALSE)

Arguments

Clustered_Dataframe	The data frame with class of "Clustered_Dataframe" which the operation will be performed on.
find_pattern	The event, event set, sequence, or a list which contains a combinations of those that are to be found.
event_set	A boolean indicator to indicate if the pattern contains the event set structure of '(event, ..., event)'
exact	a boolean indicator to indicate if the pattern should be searched for how it's exactly entered; this uses stringer 'fixed()' modifier.

Value

Returns a data frame with the columns of cluster, id, sequence, and sequences which has a class of ( "tbl_df", "tbl", "data.frame").

If specifying the event set structure, by setting event_set = TRUE, the following shorthands have been created for ease of use.

event*,	will translate into "([:alnum:]*, )*" and will match an event zero or more times. To be used when specifying any number of events are allowed to occur before a specified event.
event+,	will translate into "([:alnum:]*, )+" and will match an event one or more times. To be used when specifying at least 1 event should occur before a specified event.
, event*	will translate into "(, [:alnum:]*)*" and will match an event zero or more times. To be used when specifying any number of events are allowed to occur after a specified event.
, event+	will translate into "(, [:alnum:]*)+" and will match an event one or more times. To be used when specifying at least 1 event should occur after a specified event.
**	will translate into "[[:print:]]*" and will match any printable characters which are: [:alnum:], [:punct:] and space.
eventset*	will translate into "(\([[:alnum:], ]*[[:alnum:]*]+\) )*" and will match an event set structure zero or more times. Specifies that any number of event sets
eventset+	will translate into "(\([[:alnum:]*, ]*[[:alnum:]*]+\) )+" and will match an event set structure one or more times.
eventset*	will translate into "( \([[:alnum:], ]*[[:alnum:]*]+\))*" and will match an event set structure zero or more times.
eventset+	will translate into "( \([[:alnum:]*, ]*[[:alnum:]*]+\))+" and will match an event set structure one or more times.

Examples

data("demo1")
demo1 <- data.frame(do.call("rbind", strsplit(as.character(demo1$id.date.item), ",")))
names(demo1) <- c("id", "period", "event")


agg <- demo1 %>% aggregate_sequences(format = "%m/%d/%Y",
                                     unit = "month",
                                     n_units = 1,
                                     summary_stats = FALSE)


clustered <- agg %>% cluster_kmedoids(k = 2, use_cache = TRUE)
pats <- clustered %>% filter_pattern(threshold = 0.3, pattern_name = "consensus")


# Find all sequences which contain an event set with events K and M;
#   This allows for events to occur before, after, or between events K and M
pattern_search(pats, find_pattern = "(K, M)")
# Returns 4 sequences

# Find all sequences which contain an event set with I and M as well as
# an event set that contains L and M
pattern_search(pats, find_pattern = "(K, M), (L, M)")

# For more granular control over events within an event set, set event_set = TRUE
#   This will allow fine control over the possibility of events to occur before,
#     after, or between the events of interest

ilangurudev/approxmapR documentation built on March 22, 2022, 1:15 p.m.