Changelog • scaledic

scaledic 0.4.0

Introduced new dic attribute: recodes. Recodes is useful when you have a test item with multiple response options that are evaluated later on. You can now put the raw responses as the variable values and define recodes to tell how to evaluate the raw data.

recode_dic_items() can be used to create a new recoded variable or recode all variables in a data.frame that have recoding information in its dic-attributes.

dat <- new_dic(
  x = c(1,1,2,3,1,3,4,4,3,2,4,5),
  item_name = "knowledge_1", 
  item_label = "What is the captial of Germany?",
  type = "integer",
  weight = 1,
  values = "1:4",
  value_labels = "1 = Brussels; 2 = Hamburg; 3 = Bonn; 4 = Berlin",
  recodes = "1 = -1; 2 = 0; 3 = 0; 4 = 1"
)
dat
recode_dic_items(dat)

New function

remove_dic_class() / add_dic_class(): Removes/ adds the dic class from a vector/ variable.

scaledic 0.3.1

solved bug. extract_dic did not work when type name was in capitals (e.g. “Integer”)
alphy_table(): New argument keys_from_weights. If TRUE, tries to extract keys argument from scaledics weights parameter.

scaledic 0.3.0

New function

new_dic(): Constructor dic objects. Mainly used for apply_dic function.
factor_by_label(): turns a dic object into a factor based on the value_labels.

> dat_dic <- apply_dic(ex_scaledic_data, ex_scaledic_dic)
> table(factor_by_label(dat_dic$rel_1), dat_dic$gender)
                     
                      m f d
  Never               0 1 1
  Once a year or less 1 2 1
  A few times a year  1 1 3
  A few times a month 0 2 0
  Once a week         1 1 0
  More than once/week 4 1 0

scaledic 0.2.5

New functions

remove_dic(): Returns an object that does not inherit from “dic”. In case of a data.frame, the “dic” class will be removed from all variables. If argument remove_attributes is set TRUE, all dic attributes will be removed as well.
lookup_norms(): Turns raw scores to normscores with the help of a normtable. Added to example normtables for the ex_itrf example.

normtable <- data.frame(
  age = rep(c(6, 8, 6, 8), each = 11),
  gender = rep(c("m", "w"), each = 22),
  raw =  rep(0:10, 4),
  T = rep(c(40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 
            37, 39, 41, 43, 45, 47, 49, 51, 53, 55, 57), 2) + rep(c(0,5), each = 22)
)

rawscores <- c(5,5,3,1)
group_age <- c("6", "8", "6", "8")
group_gender <- c("m", "m", "w", "w")

lookup_norms(rawscores, group = list(age = group_age, gender = group_gender), normtable)

## When group values are not specified exactly, raw scores can be ambiguous:
lookup_norms(rawscores, group = list(gender = group_gender), normtable = normtable)

lookup_norms(rawscores, normtable = normtable)

get_scales(): A wrapper around select_items() to extract multiple scale definitions by providing logical expressions or a name:

# providing individual logical expressions
scales <- get_scales(ex_itrf,
  'APD' = subscale_2 == "APD",
  'OPP' = subscale_2 == "OPP",
  "SW" = subscale_2 == "SW",
  "AD" = subscale_2 == "AD"
)

# providing a name of a dic attribute
scales <- get_scales(ex_itrf, subscale_2)

# are identical to:

scales <- list(
  'APD' = select_items(ex_itrf, subscale_2 == "APD", names_only = TRUE),
  'OPP' = select_items(ex_itrf, subscale_2 == "OPP", names_only = TRUE),
  "SW" = select_items(ex_itrf, subscale_2 == "SW", names_only = TRUE),
  "AD" = select_items(ex_itrf, subscale_2 == "AD", names_only = TRUE)
)

get_dic_attribute(). Returns a list with dic attributes for a dataframe.

ex_itrf %>% 
  select_items(subscale == "Int") %>% 
  get_dic_attribute("item_label")

Reworked

rename_items(): Much more versatile syntax applying the tidyvers glue function. Old functionality kept but throws a deprecated warning.)

ex_itrf %>% select(1:5) %>%
  rename_item("{reverse}{item_name}: {item_label}")

scaledic 0.1.27

Bugs:
- haven_dic(): labels with length > 0 are not set (a warning is thrown).

Changes: - apply_dic(): replaces linebreaks in value_labels with “;”

scaledic 0.1.20

New functions: - set_label(): Sets dic information. Will replace set_dic.

Changes: - apply_dic(): new attribute coerce_class. When type of dic is numeric it checks if class of variable is numeric too. If not, it throws a message and coerces type to numeric. When coerce_class = FALSE, check is applied but no coercion. - dictionary file: New variable “active”. When a column “active” exists, all tows with active != 1 are dropped. - exploratory_fa(): explained variance added to output

scaledic 0.1.16 (2020-01-12)

A lot of bug fixes

scaledic 0.1.15 (2020-12-12)

implemented vctrs vec_ptype2 vec_cast S3 methods for dic variables for better integration of scaledic into dyplyr

scaledic 0.1.12 (2020-09-22)

New function: comnbine_data_frame(). Will combine several dataframes (rows and columns) into a new data frame while keeping the dic information

scaledic 0.1.11 (2020-09-22)

apply_dic(): impute_values before scoring when impute_values = TRUE.
select_scale() is deprecated. Instead use select_items(data, filter)
get_index() is deprecated. Instead use select_items(data, filter, names_only = TRUE).

scaledic 0.1.10 (2020-06-26)

Internally reworked the values and value_labels attributes.

scaledic 0.1.9 (2020-04-16)

New function: score_from_dic(). Now, a dictionary file can contain information for building scale score (score_filter, score_function, item_label, item_name. This function extracts these infromation from the dic file and returns a data frame with the new score variable(s).

scaledic 0.1.8.2 (2020-04-07)

New function: rename_by_list() for renaming variables in a data file based on a provided list with named strings.

scaledic 0.1.8.0 (2020-03-25)

New function: descriptives() for returning a table with descriptive statistics for variables in a data.frame.
apply_dic(): If a charcter string is passed to the dic argument instead of a data frame, the function will take this as a filename and tries to load the file.
get_dic() renamed to backup_dic()

scaledic 0.1.7.10

New function: alpha_table() for creating an item analyses for mutiple scales
new function: psych_fa() for returning a loading matrix of a factor analyses (based on the psych package)

scaledic 0.1.7.9

New format for specifiying selection of items by dic information: now uses logical operations including all dic attributes. Examples: select_scale(data, scale == "ITRF"), get_index(data, scale == "ITRF" & subscale = "Int" & weight == 1). The old specification is still working but deprecated. New format is applicable in: get_index(), select_scale(), impute_missing(), score_scales().
dicfile: sub_scale and sub_scale_2 renamed to subscale and subscale_2. Old dic files still work as variables are renamed to new versions on import.
Internal reorganization: dic attributes sub_scale and sub_scale_2 renamed to subscale and subscale_2.

scaledic 0.1.7.8

score_scale(): New argument sum. If TRUE, calculated the sum and if FALSE, calculates the mean. Argument label can be used to set a dic label for the resulting score variable. When label is left NULL, a label will be generated automatically.
apply_dic(): Agrument set_label_attr is now TRUE by default.

scaledic 0.1.7.7

Internal rename of dic attribute names.
Variable name of dic file is now “name” by default. label is still possible but will be renamed internally.

scaledic 0.1.7.6

New class dic with a [ method. This allows to subset dataframes with dictionaries without loosing the dic information. The class is added as the fist class to each variable in a dataframe that contains dic information.
print() method for class dic.
dic_file: new type values possible: ‘numeric’ or ‘double’
get_index() now takes any dictionary attribute and returns an error message when an invalid dic attribute is selected.
dic_haven(): Adds Hmisc/haven labels from dic infos
haven_dic(): Adds dic infos from haven labels
apply_dic(): New argument set_label_attr. If TRUE executes dic_haven().

scaledic 0.1.7

apply_dic(): Bug with factor type solved

scaledic 0.1.6

new function build_scaledic_skeleton(): Creates an Excel file with a template for a dic file.
list_scales(): new argument n_items for displaying number of items per scale.
list_scales(): new argument char_na for displaying NAs.
dic_file: new type values possible: ‘string’, ‘character’, or ‘characters’
names2labels(): new argument prefix takes character vector with possible values ‘scale’, ‘subscale’, ‘subsclae2’, ‘reverse’, and ‘weight’ adding the respective information before the item.
names2label(): deletet arguments short and reverse.
names2label(): New arguments char_weight, char_sep, and char_prefix_end help to specify look of output.
apply_dic(): new replace_missing argument replaces missing values as defined in the dic file with NAs.