Annual Daysim Lighting Profile with Manual On/Off control "0": schedule has fractional values



Hi everyone!

I was trying to create a schedule for EnergyPlus based on HB Annual Daylight simulation, having the lighting control type set to “0” for manual on/off switch. I used Mostapha’s script for Integrated Daylight and Energy from Hydra as a guideline.

My problem is that the manual on/off schedule, as I understand it, should only have values for ‘on’ = 1 and ‘off’ = 0, but I get occasional fractional values in the resulting AnnualProfile schedule (see image attached). Is this expected to happen?

I am also wondering what is the logic behind this lighting control type.
Q1: There is a sensor point but no illuminance threshold assigned, so what are the criteria for users switching lights on/off?
Q2: What is the user’s reaction time and behaviour patterns in responding to the lighting conditions?

Big thanks in advance for your help,


This comes from the way the lightswitch algorithm is designed. The algorithm assumes the following scenarios;
Lighting Control | Blind Control | User Behavior
active | active | User uses electric lighting in accordance with the need only
passive | active | User keep electric lights on and uses blinds when needed
active | passive | User keep electric lights when needed and keep the blinds closed
passive | passive | User keep electric lights on and keeps the blinds also closed

now there’s no way to know what is the probability of any one of the behavior happening at a particular hour. Therefore, all the four scenarios are considered and the average lighting energy use for the hour is defined. I believe this is the reason you are seeing floats (not whole numbers).


Thanks a lot @devang ! :slightly_smiling_face:

Now if I am not mistaken, this Daysim algorithm is based on a research done by Christoph Reinhart that can be found in his publication Lightswitch-2002: a model for manual and automated control of electric lighting and blinds. Posting it here if anyone is looking for a reference.