nhs-bnssg-analytics / fpt_analysis

Scoping the possibility of predicting performance from demand and capacity metrics
1 stars 0 forks source link

Imputing data #15

Open sebsfox opened 11 months ago

sebsfox commented 11 months ago

Should I impute missings differently? For example - bed availability overnight for Learning Disabilities for some ICSs have data (eg, a positive numerator and denominator) one year, and then none the next year (see RV5, which stops reporting in April 15 but there are no comments as to why). Should that have a value of 0 for value? Does it even matter if we are using numerators and we have a 0 for those time periods?