Fixes to various row_labels situations with tests. #700
Conversation
|
Code Coverage SummaryDiff against mainResults for commit: 54fb8e7 Minimum allowed coverage is ♻️ This comment has been updated with latest results |
There was a problem hiding this comment.
Thanks @gmbecker. Here is my feedback on some minor things:
- I think it would be better if the analysis variable and analysis function are always annotated regardless of the type of table variants {1-4} as described in [Bug]: row_labels don't match docs in qtable() #698.
If users think the default label is too long it can easily be changed withrow_labels.
For example:
qtable(ex_adsl , col_vars = "ARM")
Should return
A: Drug X B: Placebo C: Combination
(N=134) (N=134) (N=132)
——————————————————————————————————————————————————————————
STUDYID - count 134 134 132
-
What do you think about changing the default formatting of
avar - afunlabel toafun(avar). For example, instead ofAGE - maxtomax(AGE)? Given that the purpose is exploratory, this would be a nice reminder to the person doing the analysis how the table calculations are done. -
This should throw an error as the
row_labelslength shouldn't exceed 1 for table variant {2}:
qtable(ex_adsl , row_vars = "STRATA2", col_vars = "ARM", avar = "AGE", afun = mean,
row_labels = c("ABC", "EFG", "HIJ"))
ABC
EFG A: Drug X B: Placebo C: Combination
HIJ (N=134) (N=134) (N=132)
—————————————————————————————————————————————
S1 34.10 36.46 35.70
S2 33.38 34.40 35.24
- For multi-row returning afun the length of
row_labelsshould match the number of cells created byafun. Right now nothing happens but I think this should also throw an error. Same thing happens in the case of this table with the additionalrow_varssplit:
qtable(ex_adsl, col_vars = "ARM", avar = "AGE", afun = fivenum3,
row_labels = "ABC")
AGE - fivenum3 A: Drug X B: Placebo C: Combination
(N=134) (N=134) (N=132)
————————————————————————————————————————————————————————
21.00 21.00 20.00
28.00 30.00 30.00
33.00 35.00 35.00
39.00 40.00 40.00
50.00 62.00 69.00
|
@anajens I didn't change the label display but both directions of row_labels - number of afun cells mismatch are now errors. Also there are now separate |
Thanks! I think this is a good start but not exactly what I had in mind. I was thinking about being able to return the expression of the layout so users can see directly the connection from qtable to the layout framework. But that's probably a much bigger task for another day ( if users request this). |
That is possible to do but out of scope for this issue, as it would require non-trivial new development. It is a good idea though |
|
New issue added: #711 |
Close #698.