gs_power_npe and gs_design_npe enforce a restriction that the final value in theta must be positive. I can see this is likely strictly needed for gs_design_npe in order to be able to power the trial as desired, but it does not seem necessary for gs_power_npe (manually calling the code in gs_power_npe with a negative theta doesn't cause any trouble for me and I see no reason why it would).
Would it be possible to remove this restriction in gs_power_npe, as its preventing use of the function to calculate futility stopping when the effect is indeed negative?
gs_power_npeandgs_design_npeenforce a restriction that the final value inthetamust be positive. I can see this is likely strictly needed forgs_design_npein order to be able to power the trial as desired, but it does not seem necessary forgs_power_npe(manually calling the code ings_power_npewith a negativethetadoesn't cause any trouble for me and I see no reason why it would).Would it be possible to remove this restriction in
gs_power_npe, as its preventing use of the function to calculate futility stopping when the effect is indeed negative?