ELR feedback 28 June 2021

whocov / trend_analysis_public

Public version of the trend analysis infrastructure

https://asmodee-infrastructure-handbook.netlify.app/

Other

2 stars 3 forks source link

ELR feedback 28 June 2021 #25

Closed thibautjombart closed 3 years ago

thibautjombart commented 3 years ago

Here is a list of things to address for the next release

[x] ...

nsbatra commented 3 years ago

[x] It would be nice to have Excel columns in the following order: assessment date, country, iso2, region, population, growth rate, relative incidence, net increase, last_week_trp, tpr_change, epi_classification, r_upper, r_lower, r_tpr
[x] Excel export still has >400 rows, probably to do a join that isn't matching 100%. It would be nice if the join could be fixed so each country appears once only
[x] Nomenclature standardisation: in column epi_classification can "NL" become "No concern", and "Mid" become "Medium" @thibautjombart

thibautjombart commented 3 years ago

Thanks for this, very useful. I am also keen to standardise the format of columns. Are you happy with the following?

[x] decimal numbers rounded at 3 digits (as currently for r)
[x] proportions output as such, leaving units (e.g. %) down to the 'format' of the corresponding columns; this would change e.g. incidence as it is currently (and stupidly on my end) output as text

MustafaMahran commented 3 years ago

[ ] maybe automating vaccination coverage as you did TPR

thibautjombart commented 3 years ago

Thanks for this. Will discuss with @henryls1 if the data stream we use can allow it, but I agree it would be a nice addition. @nsbatra what do you think?

henryls1 commented 3 years ago

The data should allow for vaccination coverage - the only issue is that we do not have vaccine-eligable population denominators for countries, but otherwise should be trivial

nsbatra commented 3 years ago

@thibautjombart yes, agree with your column format standardisations I am also OK with trying to automate vax coverage to reduce work for intel

nsbatra commented 3 years ago

[x] Two countries remain problematic in that they are not being assigned WHO regions in the Excel export: Cote d'Ivoire and Northern Mariana Islands.

I suspect "Northern Mariana Islands" is not matching to a WHO region because its formal name is "Northern Mariana Islands (Commonwealth of the)". I am not sure why Cote D'Ivoire is not matching.

Perhaps @thibautjombart can re-construct the joins that create the Excel to use iso3 instead of report_country? I looked quickly but am not sure where in the Rmd these are

thibautjombart commented 3 years ago

@nsbatra the following has a mix of capitalisation:

It would be nice to have Excel columns in the following order: assessment date, country, iso2, region, population, growth rate, relative incidence, net increase, last_week_trp, tpr_change, epi_classification, r_upper, r_lower, r_tpr

Could you check and confirm what you want to use as final column names for the xlsx output?

thibautjombart commented 3 years ago

I am moving the bit on 'adding vaccination' to a separate issue: https://github.com/whocov/trend_analysis_public/issues/26 due to this information not being part of the current output of assemble_data.Rmd

thibautjombart commented 3 years ago

Closed via 27faffcec3e6d93442288db44d8d6547f5a84245