Grain decay appears to do nothing...

Created by: Lars Dalby

Imported comments:

By Lars Dalby on 2018-01-31T13:06:56.149Z

closed

By Lars Dalby on 2018-01-31T13:06:55.739Z

Then we get this: image

So a clear improvement to the fits. So the lack of response in the previous runs was due to a too high default for WC to grass scaler

By Lars Dalby on 2018-01-31T12:31:08.385Z

Damn, a strange bug in the plot that I can't get rid off right now. Anyway we get this (ignore what looks like 100% grain usage in for barnies in sep.):

image.

The fits are alright I think. But they are worse in this run compared to the previous one. So it looks like something is going wrong on grendel.

However the difference is not as big as you would expect. This must be due to the too high energetic value of WC. Let's set WC = grass and see what happens

By Lars Dalby on 2018-01-31T09:38:42.104Z

This is likely to be an issue where the default for winter cereal to grass scaler being too high. So not matter how much grain there is, they eat WC.

This is a default run: image Let's see what a run with a very high decay rate looks like.

By Lars Dalby on 2018-01-31T07:59:49.191Z

marked this issue as related to #23 (closed)

By Lars Dalby on 2018-01-31T07:57:05.475Z

This is running with a WC to grass scaler that should give a low fit (according to the latest param fit document where I believe we haven't subtracted 1 from the lsq) The fits shown here is when I subtract 1 from the fit! image

And this is with a value that should give a high fit: image

Have trawled the code to see if there is some odd place where the subtraction is going on, but that doesn't seem to be the case. Based on this and the plots above I think it is safe to conclude that it is not being done and should be.

By Lars Dalby on 2018-01-30T20:36:22.001Z

Looks like this is switched around, so lower values for habuse fit is better, but still not showing any change which it should.

By Lars Dalby on 2018-01-30T15:21:32.331Z

Have a better guess at the FL default now, so proceeding here.

By Lars Dalby on 2018-01-29T15:49:04.793Z

We need a new sensible default for FL before we can really fix this

By Lars Dalby on 2018-01-29T13:56:18.759Z

added ~3117938 and removed ~3117937 labels

By Lars Dalby on 2018-01-29T13:45:42.243Z

image

Those values corresponds to these cfg values: 0.0995, 0.2786, 0.4577, 0.6368, 0.8159, 0.9950. This means that with the lowest of those, all grain is pretty much gone by day 3. This should really give us a response.

By Lars Dalby on 2018-01-08T11:00:16.075Z

added ~3117941 label

By Lars Dalby on 2018-01-08T11:00:07.829Z

assigned to @LDalby