expss (version 0.5.5)

dichotomy: Convert variable (possibly multiple choice question) to matrix of dummy variables.

Description

This function converts variable/multiple response variable(matrix/data.frame) with category encoding into matrix with dichotomy encoding (0/1) suited for most statistical analysis, e. g. clustering, factor analysis, linear regression and so on.

• dichotomy returns matrix with 0, 1 and possibly NA.
• dichotomy1 drops last column in dichotomy matrix. It is useful in many cases because any column of such matrix usually is linear combinations of other columns.
• dummy is another shortcut for dichotomy.
• *_df are the same functions as dichotomy etc. but return data.frame instead of matrix.

Usage

dichotomy(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)
dichotomy1(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)
dichotomy1_df(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)
dichotomy_df(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)
dummy(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)
dummy_df(x, prefix = NULL, keep_unused = FALSE, use_na = TRUE, keep_values = NULL, keep_labels = NULL, drop_values = NULL, drop_labels = NULL)

Arguments

x

vector/factor/matrix/data.frame.

prefix

character. If it is not NULL it instead of labels will be used prefix+values.

keep_unused

Logical. Should we create columns for unused value labels/factor levels.

use_na

Logical. Should we use NA for rows with all NA or use 0's instead.

keep_values

Numeric/character. Values that should be kept. By default all values will be kept.

keep_labels

Numeric/character. Labels/levels that should be kept. By default all labels/levels will be kept.

drop_values

Numeric/character. Values that should be dropped. By default all values will be kept. Ignored if keep_values/keep_labels are provided.

drop_labels

Numeric/character. Labels/levels that should be dropped. By default all labels/levels will be kept. Ignored if keep_values/keep_labels are provided.

Value

matrix or data.frame with 0,1 which column names are value labels or values with prefix. If label doesn't exist for particular value then this value will be used as column name.

See Also

category for reverse conversion.

Examples

# toy example
# brands - multiple response question
# Which brands do you use during last three months? 
set.seed(123)
brands = t(replicate(20,sample(c(1:5,NA),4,replace = FALSE)))
# score - evaluation of tested product
score = sample(-1:1,20,replace = TRUE)
var_lab(brands) = "Used brands"
val_lab(brands) = make_labels("
                              1 Brand A
                              2 Brand B
                              3 Brand C
                              4 Brand D
                              5 Brand E
                              ")

var_lab(score) = "Evaluation of tested brand"
val_lab(score) = make_labels("
                             -1 Dislike it
                              0 So-so
                              1 Like it    
                             ")

# percentage of used brands
colMeans(dichotomy(brands))

# percentage of brands within each score
cro_mean(dichotomy(brands), score)
# the same as
cro_cpct(brands, score)

# percentage of brands within each score - same numbers
aggregate(dichotomy(brands) ~ f(score), FUN = mean)

# or, same result in another form
by(dichotomy(brands), f(score), FUN = colMeans)

# customer segmentation by used brands
kmeans(dichotomy(brands),3)

# model of influence of used brands on evaluation of tested product 
summary(lm(score ~ dichotomy(brands)))

# prefixed data.frame 
dichotomy_df(brands, prefix = "brand_")