- Shading for Confidence Intervals in
visualize()
- Need to implement for theoretical distributions and
method = "both"
- Need to implement for theoretical distributions and
- Allow
visualize()
to specify whichvariable
to plot instead of only working withstat
column - Check that corresponding z for one prop depends on
params
being set inhypothesize
withspecify() %>% calculate()
shortcut - Check that assumptions have been met for the theoretical distribution and
"both"
- Fix double printing of
Response:
andExplanatory:
- Might have something to do with
infer
class being set in multiple spots?
- Might have something to do with
- Check that
specify() %>% calculate()
works forstat = "z"
andstat = "t"
- Determine if other wrapper functions should be created
z_test()
,mean_stat()
,diff_in_mean_stat()
, etc.?
- Determine if other wrapper functions should be created
- Consider re-working how p-values can be calculated (both for computational and theoretical)
- Maybe check to make sure that p-values don't exceed 1 too
- Should the wrapper functions like
t_test()
also include a logicalconf_int
argument? - Check that
stat
is calculated appropriately ifgenerate()
is not calledorder
of the arguments into the standardized statistics should be checked again- May require moving of
set_params()
to be afterspecify()
- Update
visualize()
tests to includecalculate()
step formethod = "theoretical"
- Add
add_obs_stat
toggle intovisualize()
and p-value calculation? - Shift to list-columns in
generate()
- Implement check of
stat
in theoreticalvisualize()
- Write test to check that bootstrapped values are centered near the
hypothesized value in
specify()
- Write vignettes on how NOT to use infer (strange errors, funky results, etc.)
- Implement theoretical distributions for bootstrap distributions
- Create
resources.md
with links to slides/talks/workshops given aboutinfer
- Add Lionel's vis checking package (
vdiffr
) tovisualize()
tests - Determine whether
calculate()
should be where theset_params()
function is called instead of inspecify()
- Need to also add parameters to wrapper functions so that randomization methods can be implemented by practitioners looking to skip the longer pipe syntax
- Find a better word than
"simulate"
fortype
ingenerate()