replace cases by confirm for API consistency

[avallecam]

Wondering if for consistency across the ecosystem, {cfr} and {EpiNow2} should use a similar column name. Data from {cfr} can also be used for {EpiNow2} right?

library(EpiNow2)
library(cfr)

head(EpiNow2::example_confirmed)
#>          date confirm
#>        <Date>   <num>
#> 1: 2020-02-22      14
#> 2: 2020-02-23      62
#> 3: 2020-02-24      53
#> 4: 2020-02-25      97
#> 5: 2020-02-26      93
#> 6: 2020-02-27      78
head(cfr::ebola1976)
#>         date cases deaths
#> 1 1976-08-25     1      0
#> 2 1976-08-26     0      0
#> 3 1976-08-27     0      0
#> 4 1976-08-28     0      0
#> 5 1976-08-29     0      0
#> 6 1976-08-30     0      0

^{Created on 2024-03-19 with reprex v2.1.0}

[pratikunterwegs]

In principle having common names makes sense. I think {EpiNow2} should change its requirements for column names as 'cases' alongside 'deaths' is IMO clearer to users; 'confirm' is ambiguous. In practice, changing requirements in either package would be a breaking change, and therefore difficult to justify for uniformity's sake.

[sbfnk]

See epiforecasts/EpiNow2#505

[pratikunterwegs]

Thanks, I hadn't seen that issue! Again, in practice, if we do decide to standardise, it does make make sense for {cfr} to change its requirement to match {EpiNow2} - if the new name makes sense - as there are far fewer {cfr} users (I imagine) so lower impact overall.

[pratikunterwegs]

Hi @avallecam - could I get an idea if this issue is still relevant or if it can be closed?

[avallecam]

Hi @avallecam - could I get an idea if this issue is still relevant or if it can be closed?

I'll look at this at the end of this week again.

[avallecam]

Hi @avallecam - could I get an idea if this issue is still relevant or if it can be closed?

May {cfr} and {EpiNow2} be facing a similar situation? the count column of their input data could contain many types of observation, always linked with the corresponding delay.

• {cfr} from observation to death,
• {EpiNow2} from infection to observation.

Following Gosnic et al., Fig 4, observation can be any "observed data" between infection and death.

The column name of cases for {cfr} and confirm for {EpiNow2} are defined given the most immediate and memorable need that each of these packages satisfy. But they reduce the visibility of alternative observation-delay combinations.

From a learning point of view:

• For the novice learner, keeping the names as they are minimize the effort to learn.
• Then, for the competent practicioner, we can expect them to rename columns to the type of observation they have available (as mentioned in this EpiNow2 comment)

---

I suggest to keep this issue open and wait for {EpiNow2} to see how they generalize the column name and evaluate to replicate in {cfr}. A similar column name may facilitate to visualize the relationship and flexibility of these two packages.

In the meantime, we may want to increase the visibility of the alternative observation-delay combinations. For {cfr}, IFR and HFR are only mentioned at the end of the first vignette. This can be in an exclusive vignette or howto entry. As a reference, we have a PR to include how to reconstruct the infection curve from deaths using {EpiNow2}.

[pratikunterwegs]

Thanks @avallecam, I'll keep this issue open for now.