Note about "no visible binding"

[eduardgrebe]

Status: 1 NOTE
checking R code for possible problems ... NOTE
fit_binomial_model: no visible binding for global variable
  ‘tolerance_glm2’
fit_binomial_model: no visible binding for global variable ‘maxit_glm2’
frrcal: no visible binding for global variable ‘time_since_eddi’
frrcal: no visible binding for global variable ‘sid’
integrate_for_mdri: no visible binding for global variable
  ‘tolerance_integral’
integrate_for_mdri: no visible binding for global variable
  ‘maxit_integral’
mdrical: no visible global function definition for ‘%dopar%’
mdrical: no visible binding for global variable ‘sid’
plot_probability: no visible binding for global variable
  ‘time_since_eddi’
plot_probability: no visible binding for global variable ‘probability’

[eduardgrebe]

Down to

Status: 1 NOTE
checking R code for possible problems ... NOTE
frrcal: no visible binding for global variable ‘time_since_eddi’
frrcal: no visible binding for global variable ‘sid’
mdrical: no visible global function definition for ‘%dopar%’
mdrical: no visible binding for global variable ‘sid’
plot_probability: no visible binding for global variable
  ‘time_since_eddi’
plot_probability: no visible binding for global variable ‘probability’

[eduardgrebe]

I can see no way of resolving this note at present. 'time_since_eddi', 'sid' and 'probability' are names assigned to columns in dataframes which are later referred to without the '$' notation because of the way the functions in question work. So it erroneously believes these to be global variables, when they are in fact not.

The issue with '%dopar%' is that doMC and foreach are not imported. They are only installed and loaded on non-Windows systems. If we made them dependencies the package would become incompatible with Windows. I haven't actually tried importing them and seeing what it does on Windows. I might check that now.

[eduardgrebe]

Probably not fixable by v1.0.0 milestone.

[eduardgrebe]

This is actually fixed now.