You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Say user specifies stratified PS models for A (resp. C or N) nodes such that the union of the person-time observations on A that fall in each stratum is not equal to the total number of person-time observations in the dataset.
Say that the user then asks stremr to compute an effect estimate for interventions on a subset of the A nodes (resp. C or N) such some elements of the subsets of person-time observations on A do not fall into any of the strata specified for gA.
Stremr should then return an error message to warn the user than they did not specify a sufficient number of strata for gA. Currently, stremr will return an output which will be incorrect.
Idea to fix this: add an argument check in getIPWeights that checks that the A (resp. C and N) person-time observations on which the user wish to intervene are all contained in the union of the person-time observations on A that fall within each stratum specified for gA. If not, stremr should give an error that asks the user to modify/add how they specified the strata for the PS model for gA.
The text was updated successfully, but these errors were encountered:
Say user specifies stratified PS models for A (resp. C or N) nodes such that the union of the person-time observations on A that fall in each stratum is not equal to the total number of person-time observations in the dataset.
Say that the user then asks stremr to compute an effect estimate for interventions on a subset of the A nodes (resp. C or N) such some elements of the subsets of person-time observations on A do not fall into any of the strata specified for gA.
Stremr should then return an error message to warn the user than they did not specify a sufficient number of strata for gA. Currently, stremr will return an output which will be incorrect.
Idea to fix this: add an argument check in getIPWeights that checks that the A (resp. C and N) person-time observations on which the user wish to intervene are all contained in the union of the person-time observations on A that fall within each stratum specified for gA. If not, stremr should give an error that asks the user to modify/add how they specified the strata for the PS model for gA.
The text was updated successfully, but these errors were encountered: