search

msr-consulting / exscalabar_server

Repository for the EXSCALABAR server.
http://www.msrconsults.com/ukmet-gh/exscalabar
0 stars 1 forks source link

Vaisala cal coefficients #161

Closed JustinLangridge closed 7 years ago

JustinLangridge commented 7 years ago

Can we add calibration coefficients for the vaisalas to the .ini file? These will be coefficients based on a straight line fit of the true RH to the RH measured by the device.

While this is a nicety for most of the probes and can be applied in post-processing, it is important for the two vaisalas used to control the RH as we want to control to a calibrated RH value.

lo-co commented 7 years ago

Providing two new properties in the Configurable Device class:

The equation that is put in for the Cal Equation is a polynomial that looks like

[ID].CalEquation = Cn*x^n+Cn-1*x^(n-1)+...+C1*x+C0

where x is the original value and Cn is the nth coefficient. This value is optional in the INI file and the default is 1*x.

lo-co commented 7 years ago

Added for MFCs, PPTs and Vaisalas. Tested specifically for Vaisala's.

datid commented 7 years ago

Is this calibration just for display purposes ? I think data recorded should be as close to raw data as possible.

lo-co commented 7 years ago

No. It is not just for display purposes. Generally calibrations are generally something done rarely and therefore we just apply them across the board. The raw data is always there - the user performing the analysis simply has to invert the calibration (generally trivial for polynomial fits). That being said, I understand the desire to include the raw data. To this end, I have two potential fixes (neither of which I will include in this phase as they are beyond the scope of work - but you are always welcome to change them):

Personally, I think that the second is more elegant - this is what we generally do for calibrations of measurements such as flow, T and RH. But, either one is workable (the first probably requires the most work).

datid commented 7 years ago

Seems the second is probably the better option.

From: Matt Richardson [mailto:notifications@github.com] Sent: 30 November 2016 05:41 To: msr-consulting/exscalabar_server Cc: Tiddeman, Dave; Comment Subject: Re: [msr-consulting/exscalabar_server] Vaisala cal coefficients (#161)

No. It is not just for display purposes. Generally calibrations are generally something done rarely and therefore we just apply them across the board. The raw data is always there - the user performing the analysis simply has to invert the calibration (generally trivial for polynomial fits). That being said, I understand the desire to include the raw data. To this end, I have two potential fixes (neither of which I will include in this phase as they are beyond the scope of work - but you are always welcome to change them):

Personally, I think that the second is more elegant - this is what we generally do for calibrations of measurements such as flow, T and RH. But, either one is workable (the first probably requires the most work).

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/msr-consulting/exscalabar_server/issues/161#issuecomment-263788217, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ALmnR5tBb3MoZ6rxwkdEwwnwEtLOJYXMks5rDQxOgaJpZM4K5sEv.

lo-co commented 7 years ago

If it is quick I might be able to get it in. I think it would be good to always have the most recent copy of the INI anyway just in case something happens.