Closed ttngu207 closed 1 week ago
@glopesdev it seems like this is probably a bug where the PatchState is updating before the BlockState. In what cases can this occur?
And I guess for analysis for social02, we're fine just keeping the first rate (in blocks with >1 rate) as the "true" rate
Is this a bug? At the block transition we don't change the rate until the next pellet delivery.
The csv only has the block times. Can you give an example the list of the rates in one block?
Is this a bug?
Yes.
Can you give an example the list of the rates in one block?
One patch in one block will have multiple rates: e.g. "100", "300". The bug seems to be that the BlockState file somehow gets written to twice around the time of a block transition, instead of just once, probably due to a race condition
These double 0's still occur in BlockState, but guess the second 0 can just kind of be safely ignored? @glopesdev
These double 0's still occur in BlockState, but guess the second 0 can just kind of be safely ignored?
yes.
In Block level analysis - we expect only one patch rate and patch offset for each patch.
However, in some blocks, there are multiple values found. We need to determine which value to use from the multiple values.
See attached csv file for blocks with the errors
df.csv