scale value is the critical parameter used to control the linear width for biexp transformation.
It is currently stored in various places of diva xml and often time they don't agree with one another, so it is important to choose the right version of its value.
Here are the 5 different locations where you can parse the scale_value from the same xml
gate level within tube (.../tube/gates/gate)
We can pretty much ignore experiment level since it typically stores -1. Based on the testing, Global Sheet version isn't always accurate. So we are currently only considering between the last two: tube vs gate(within tube).
However It's not clear which one is actually used by diva. Technically speaking, the same channel can be potentially scaled differently in different gates/plots since diva doesn't scale the data until it is displayed. So gate-specific scale value makes sense and we used it by default and didn't see any issues until the parser failed on a use case recently. By switching to tube level scale value (with a hidden option scale_level = "tube"), the issue resolved.
This prompt me to change the default to tube level ( this hidden option is the product of uncertainty behavior of the diva parser and not intended to be used by users).
So far, tube-level scaling passes all the existing tests and we are yet to see if there will be a gate-level scaling use case in future(I haven't been able to produce one with diva software so far)
scalevalue is the critical parameter used to control thelinear widthforbiexptransformation. It is currently stored in various places of diva xml and often time they don't agree with one another, so it is important to choose the right version of its value. Here are the 5 different locations where you can parse thescale_valuefrom the same xml/experiment/instrument_settings/parameter/)`/experiment/../worksheet_template/instrument_settings/parameter/).../worksheet_template/gates/gate)/experiment/specimen/tube/instrument_settings/parameter/).../tube/gates/gate) We can pretty much ignore experiment level since it typically stores-1. Based on the testing, Global Sheet version isn't always accurate. So we are currently only considering between the last two: tube vs gate(within tube).However It's not clear which one is actually used by diva. Technically speaking, the same channel can be potentially scaled differently in different gates/plots since diva doesn't scale the data until it is displayed. So gate-specific
scalevalue makes sense and we used it by default and didn't see any issues until the parser failed on a use case recently. By switching totubelevel scale value (with a hidden optionscale_level = "tube"), the issue resolved. This prompt me to change the default totubelevel ( this hidden option is the product of uncertainty behavior of the diva parser and not intended to be used by users). So far, tube-level scaling passes all the existing tests and we are yet to see if there will be a gate-level scaling use case in future(I haven't been able to produce one with diva software so far)