check_study_data_frame: Check various characteristics of the study data frame before...
In wfforrest/maeve: Modeling Accessories Enabling Visualization of Experiments

Description Usage Arguments Value Author(s) References Examples

Check various characteristics of the study data frame before fitting longitudinal models.

check_study_data_frame(
  input_data_frame,
  study_id_name = NULL,
  studyProtocol_name = NULL,
  return_all_columns = FALSE,
  converted_factor_order = c("observed", "alphanumeric"),
  order_rows = TRUE,
  freeze_factor = TRUE,
  autoconvert_types = TRUE,
  print_warnings = TRUE,
  return_unique = TRUE,
  group_name = maeve_options("group_name"),
  subject_ID = maeve_options("subject_ID"),
  x_name = maeve_options("x_name"),
  endpoint_name = maeve_options("endpoint_name"),
  reference_Dunnett = maeve_options("reference_Dunnett"),
  progress = maeve_options("progress"),
  abbreviate_n = maeve_options("abbreviate_n")
)

`input_data_frame`	a data.frame with raw data from DIVOS.
`study_id_name`	character string with name of study (usually a 4- or 5-digit number).
`studyProtocol_name`	character string with name of study protocol (e.g., a LASAR protocol).
`return_all_columns`	logical: return all the columns in the processed data.frame, or just a minimally required subset?
`converted_factor_order`	character string: either 'observed' or 'alphanumeric' for whether fields (usually numeric or character) should be leveled in order of values observed or by unique alphanumeric value.
`order_rows`	logical: Should rows in returned data.frame be ordered by group, then ID, then time?
`freeze_factor`	logical: Should factors levels be set to the order observed in the returned data.frame?
`autoconvert_types`	logical. Automatically convert (1) "character" for group_name to "factor", (2) "integer" or "numeric" for subject_ID to factor(character()), (3) "integer" for x_name to "numeric"?
`print_warnings`	logical. Print warnings if automatic conversions are taking place?
`return_unique`	logical. Return only the unique rows based on the columns extracted?
`group_name`	character column name for the group name factor
`subject_ID`	character column name for the subject name factor
`x_name`	character column name for the x-axis / time field
`endpoint_name`	character column name for the x-axis / time field
`reference_Dunnett`	character string with the group factor level to use as Dunnett reference. If NULL, the first level should be used.
`progress`	logical. Print out various progress messages from the functions.
`abbreviate_n`	numeric. Maximum number of characters in a group name.

An R data.frame

Bill Forrest <forrest@gene.com>

Bill Forrest forrest@gene.com

www.r-project.org

   cat('exmaples in dontrun{} code block.')
   ## Not run: 
   ### (1) Will generate a (typically harmless) warning that 'x' is class 'integer',
   ###     but converted to 'numeric':
   fake_data_01 = expand.grid( ID = paste0('ID_0',1:6) , x = as.integer( 1:5 ) )
   fake_data_01 = dplyr::arrange( fake_data_01 , ID )
   fake_data_01 = cbind( fake_study_id = 'fake',
                         grp = rep( c('grp_01','grp_02'), each = 15 ), 
                         fake_data_01, 
                         resp = rnorm(30), 
                         dreck = 'drop_this!' )
   checked_data_01 = check_study_data_frame( 
                         fake_data_01, 
                         group_name = 'grp', 
                         subject_ID = 'ID', 
                         x_name = 'x', 
                         endpoint_name = 'resp', 
                         study_id_name = 'fake_study_id' )

   ### (2) Avoid the warning by creating 'x' as numeric instead of integer:
   fake_data_02 = fake_data_01;
   fake_data_02$x <- as.numeric( fake_data_02$x ) # explicitly set to 'numeric'
   checked_data_02 = check_study_data_frame( 
                         fake_data_02,
                         group_name = 'grp',
                         subject_ID = 'ID',
                         x_name = 'x',
                         endpoint_name = 'resp',
                         study_id_name = 'fake_study_id' )
 
## End(Not run)