library(PatientProfiles)
#> Warning: package 'PatientProfiles' was built under R version 4.2.3
plotCharacteristics
can plot the
summariseCharacteristics results, with boxplot or barplot. First we
create a mock data.
<- dplyr::tibble(
observation_period observation_period_id = c(1, 2, 3),
person_id = c(1, 2, 3),
observation_period_start_date = as.Date(c(
"1985-01-01", "1989-04-29", "1974-12-03"
)),observation_period_end_date = as.Date(c(
"2011-03-04", "2022-03-14", "2023-07-10"
)),period_type_concept_id = 0
)<- dplyr::tibble(
dus_cohort cohort_definition_id = c(1, 1, 1, 2),
subject_id = c(1, 1, 2, 3),
cohort_start_date = as.Date(c(
"1990-04-19", "1991-04-19", "2010-11-14", "2000-05-25"
)),cohort_end_date = as.Date(c(
"1990-04-19", "1991-04-19", "2010-11-14", "2000-05-25"
))
)<- dplyr::tibble(
comorbidities cohort_definition_id = c(1, 2, 2, 1),
subject_id = c(1, 1, 3, 3),
cohort_start_date = as.Date(c(
"1990-01-01", "1990-06-01", "2000-01-01", "2000-06-01"
)),cohort_end_date = as.Date(c(
"1990-01-01", "1990-06-01", "2000-01-01", "2000-06-01"
))
)<- dplyr::tibble(
medication cohort_definition_id = c(1, 1, 2, 1),
subject_id = c(1, 1, 2, 3),
cohort_start_date = as.Date(c(
"1990-02-01", "1990-08-01", "2009-01-01", "1995-06-01"
)),cohort_end_date = as.Date(c(
"1990-02-01", "1990-08-01", "2009-01-01", "1995-06-01"
))
)<- dplyr::tibble(
emptyCohort cohort_definition_id = numeric(),
subject_id = numeric(),
cohort_start_date = as.Date(character()),
cohort_end_date = as.Date(character())
)<- mockPatientProfiles(
cdm dus_cohort = dus_cohort, cohort1 = emptyCohort,
cohort2 = emptyCohort, observation_period = observation_period,
comorbidities = comorbidities, medication = medication
)
$dus_cohort <- omopgenerics::newCohortTable(
cdmtable = cdm$dus_cohort, cohortSetRef = dplyr::tibble(
cohort_definition_id = c(1, 2), cohort_name = c("exposed", "unexposed")
)
)$comorbidities <- omopgenerics::newCohortTable(
cdmtable = cdm$comorbidities, cohortSetRef = dplyr::tibble(
cohort_definition_id = c(1, 2), cohort_name = c("covid", "headache")
)
)$medication <- omopgenerics::newCohortTable(
cdmtable = cdm$medication,
cohortSetRef = dplyr::tibble(
cohort_definition_id = c(1, 2, 3),
cohort_name = c("acetaminophen", "ibuprophen", "naloxone")
),cohortAttritionRef = NULL
)<- summariseCharacteristics(
characteristicsResult $dus_cohort,
cdmcohortIntersect = list(
"Medications" = list(
targetCohortTable = "medication", value = "flag", window = c(-365, 0)
), "Comorbidities" = list(
targetCohortTable = "comorbidities", value = "flag", window = c(-Inf, 0)
)
)
)#> Warning: `summariseCharacteristics()` was deprecated in PatientProfiles 0.8.0.
#> ℹ Please use `CohortCharacteristics::summariseCharacteristics()` instead.
#> This warning is displayed once every 8 hours.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.
#> ℹ adding demographics columns
#> ℹ adding cohort intersect columns for table: medication
#> ℹ adding cohort intersect columns for table: comorbidities
#> ℹ summarising data
#> ℹ The following estimates will be computed:
#> • flag_variable_00006_variable_00005: count, percentage
#> • flag_variable_00007_variable_00005: count, percentage
#> • flag_variable_00008_variable_00005: count, percentage
#> • flag_variable_00010_variable_00009: count, percentage
#> • flag_variable_00011_variable_00009: count, percentage
#> • variable_00001: count, percentage
#> • cohort_start_date: min, q05, q25, median, q75, q95, max
#> • cohort_end_date: min, q05, q25, median, q75, q95, max
#> • variable_00003: min, q05, q25, median, q75, q95, max, mean, sd
#> • variable_00004: min, q05, q25, median, q75, q95, max, mean, sd
#> • variable_00002: min, q05, q25, median, q75, q95, max, mean, sd
#> ! Table is collected to memory as not all requested estimates are supported on
#> the database side
#> → Start summary of data, at 2024-04-16 10:31:44
#> Warning: There were 2 warnings in `dplyr::summarise()`.
#> The first warning was:
#> ℹ In argument: `dplyr::across(...)`.
#> ℹ In group 2: `strata_id = 2`.
#> Caused by warning in `base::min()`:
#> ! no non-missing arguments to min; returning Inf
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 1 remaining warning.
#> ✔ Summary finished, at 2024-04-16 10:31:45
#> ✔ summariseCharacteristics finished!
Now we show barplot example by setting plotStyle = “barplot” in
plotCharacteristics
. Similar to previous function, user can
define axes, facetVars and colorVars. Barplot shows both percentage and
count, user can filter to only percentage, or plot both.
plotCharacteristics(
data = characteristicsResult %>% dplyr::filter(estimate_type == "percentage"),
xAxis = "estimate_value",
yAxis = "variable_name",
plotStyle = "barplot",
facetVarX = "group_level",
facetVarY = NULL,
colorVars = c("variable_level")
)#> Warning in plotfunction(data, xAxis, yAxis, plotStyle = plotStyle, facetVarX, :
#> facetVarY put as NULL, plot overall
plotCharacteristics(
data = characteristicsResult,
xAxis = "estimate_value",
yAxis = "variable_name",
plotStyle = "barplot",
facetVarX = "group_level",
facetVarY = "estimate_type",
colorVars = c("variable_level"),
vertical_x = TRUE
)#> Warning: Removed 9 rows containing missing values (`position_stack()`).
#> Warning: Removed 1 rows containing missing values (`geom_col()`).
User can plot boxplot based on q25 q75 median min max in data, using
plotCharacteristics
and set plotStyle = “boxplot”. Boxplot
will be horizontal if the xAxis is set to estimate_value, and vertical
if yAxis is set to estimate_value. But for all the plots, at least one
of the xAxis or yAxis has to be estimate_value.
plotCharacteristics(
data = characteristicsResult,
xAxis = "estimate_value",
yAxis = "variable_name",
plotStyle = "boxplot",
facetVarX = "variable_name",
colorVars = c("group_level"),
vertical_x = TRUE
)#> Warning in plotfunction(data, xAxis, yAxis, plotStyle = plotStyle, facetVarX, :
#> facetVarY put as NULL, plot overall
#> Warning: Removed 2 rows containing missing values (`geom_segment()`).
#> Warning: Removed 1 rows containing missing values (`geom_segment()`).