consider adding a function that creates a npi_results object and flattens too

[parmsam]

Maybe something like this that performs both actions together:

npi_flat_search <- function( number = NULL,
enumeration_type = NULL,
taxonomy_description = NULL,
first_name = NULL,
last_name = NULL,
use_first_name_alias = NULL,
organization_name = NULL,
address_purpose = NULL,
city = NULL,
state = NULL,
postal_code = NULL,
country_code = NULL,
limit = 10L){
npi_results_obj <- npi::npi_search(number=number, enumeration_type=enumeration_type, taxonomy_description=taxonomy_description,
first_name=first_name, last_name=last_name, use_first_name_alias=use_first_name_alias,
organization_name=organization_name, address_purpose=address_purpose, city=city,
state=state, postal_code=postal_code, country_code=country_code, limit=limit)
npi::npi_flatten(npi_results_obj)
}

[frankfarach]

Thanks for the suggestion, @parmsam. Rather than add a new function, I would prefer to add a flatten argument to npi_search() itself:

npi_search(city = "New York City", flatten = TRUE)

If so, I think the default should be FALSE. I would still want to keep npi_flatten() of course. What do you think?

Just so you're aware, I'm going to be focused on responding to the rOpenSci peer review over the next couple of weeks. There are likely to be a lot of changes coming to the package, especially the documentation and error handling for npi_search().

[parmsam]

That sounds like a great idea. Got it. The rOpenSci peer review is more important. I can take a stab at adding that argument in the next week or two.

On a side note, let me know if you need any help with the documentation for the review. Feel free to assign me to a few of them.