suddengains 0.4.4
- add function 
plot_sg_intervals() to visualise session
to session intervals that are analysed for sudden gains 
- minor changes to integrate updates in tibble package
 
suddengains 0.4.3
- Suppress message about vector length from
extract_values() in the functions
create_bysg() and create_byperson() 
- Update ggplot2 functions with “deprecated” warning message
 
- Add clearer detailed output message in 
check_interval()
function 
suddengains 0.4.2
- Add option to specify the critical value for criterion three using
the argument 
sg_crit3_critical_value 
- Add option to extract values from multiple measures by specifting a
list in the 
extract_var_list of the function
extract_values() 
- Internal changes to 
plot_sg() function to work properly
with shiny 
- Make output from 
describe_sg() function clearer 
- Create shiny app illustrating main functions of this version,
available at https://milanwiedemann.shinyapps.io/shinygains/
 
- Correct typos
 
- Add funders to documentation
 
suddengains 0.4.0
- Fix error in formula used to define a cutoff value for the first
sudden gains criteria based on the Reliable Change Index (RCI; Jacobson
& Truax, 1991) in the function
define_crit1_cutoff() 
- Add formulas used to calculate standard error of measurement,
standard error of difference, and relaible change value to help file of
define_crit1_cutoff() 
- Simplify 
define_crit1_cutoff() function arguments 
- Add links to functions of other R packages to help files
 
- Add links to functions of suddengains package to help files
 
- Add option to not adjust critical value in third sudden gains
criterion
 
suddengains 0.3.1
- Make compaible with tidyr v1.0.0
 
- Fix error in 
extract_values() function 
suddengains 0.3.0
- Add function 
check_interval() to check if a given
interval is a sudden gain/loss 
suddengains 0.2.1
- Fix error message when loading SPSS datasets with the R package
foreign
 
- Calculate critical values from the student t distribution for third
criterion using 
base::qt() 
- Add functionality to change significance level (alpha) for computing
two-tailed critical value based on the student t distribution using the
argument 
sg_crit3_alpha 
suddengains 0.2.0
- Add new plot function 
plot_sg_trajectories() to plot
trajectories of individual cases 
- Add more specific warning and error messages:
- Message notification if no gains/losses are identified using:
identify_sg() or identify_sl()
 
- Error if no gains/losses are identified using:
create_bysg(), create_byperson() 
 
- Add colour option to 
plot_sg() function for different
groups 
- Fix calculation of start and end values in 
plot_sg()
function for bysg data sets 
- Update help files
- Fix spelling mistakes
 
- Clarify some arguments
 
- Add table explaining “pattern” method to 
select_cases()
function 
 
suddengains 0.1.0
First CRAN submission.