Following from Nick's review here and discussion in our meeting on 25/02/26, we now need to actually implement the planned changes to the templater (and subsequent translator updates as needed to accommodate these).
Broadly, the approach we settled on for updating the templater follows a new paradigm where rather than mapping IASR table data into IASR-formatted summary tables, the templater outputs will be designed for clarity, ease of use, standardisation and noise reduction (see above linked PR for detailed description).
We also have a set of data validation and missing/nan data handling practices that need to be implemented (see this PR) and we will be updating the templater to target the 2026 ISP tables, so the plan is to take a parallel implementation approach.
To break up this work we've chosen (at this stage) to chunk the work up by very high-level categories:
And the high-level todos are:
Following from Nick's review here and discussion in our meeting on 25/02/26, we now need to actually implement the planned changes to the
templater(and subsequenttranslatorupdates as needed to accommodate these).Broadly, the approach we settled on for updating the
templaterfollows a new paradigm where rather than mapping IASR table data into IASR-formatted summary tables, thetemplateroutputs will be designed for clarity, ease of use, standardisation and noise reduction (see above linked PR for detailed description).We also have a set of data validation and missing/nan data handling practices that need to be implemented (see this PR) and we will be updating the
templaterto target the 2026 ISP tables, so the plan is to take a parallel implementation approach.To break up this work we've chosen (at this stage) to chunk the work up by very high-level categories:
And the high-level todos are:
templaterfunctions to meet themtranslatorfunctions to take newly structured input tablestemplaterparadigm and functions