correlation {lifelogr}

Correlation between each variable vs each measure

Description

Prints and returns Pearson's correlation between each variable and each measure listed. Can pass in a dataset from create_dataset, or function calls create_dataset itself.

Usage

correlation(dataset = NA, person, variables, measures, time_var = NA)

Arguments

dataset

dataset from create_dataset that contains all variables and measures of interest

person

an instantiated Person object

variables

list of variables in person of interest, with structure list(source1 = c(var1, var2), source2 = c(var3, var4)) where source is a source of data as defined in a Person object, and var1 and var2 are variables from source1, while var3 and var4 are variables from source2

measures

list of measures in person of interest, with structure list(source1 = c(var1, var2), source2 = c(var3, var4)) where source is a source of data as defined in a Person object, and var1 and var2 are variables from source1, while var3 and var4 are variables from source2

time_var

the time variable that variables and measures are observed in (time, date, or datetime) - only needed if dataset is not passed in

Value

Pearson's correlation between each variable and each measure

Note

'correlation' uses "pairwise.complete.obs", which only computes the correlation between all complete pairs of observations.

Examples

data(EX)

dataset <- create_dataset(person = EX, 
            all_variables = list("fitbit_daily" = c("sleepDuration", 
                                                    "steps")),
            time_var = c("date"))
                      
correlation_df <- correlation(dataset, person = EX, 
                            variables = list("fitbit_daily" = 
                                                 c("sleepDuration")),
                            measures = list("fitbit_daily" = c("steps")),
                           time_var = "date")