Help req'd: BtSsm/MegaLogViewer Data Filter creation.

[bigBADbenny]

BtSsm/MegaLogViewer Data Filter creation.
I'd like to leverage the power of MegaLogViewer to get more analysis from my logs, especially as related to Knock, FKC and FLKC.
The scattergraph function is awesome but I need to filter out dat I don't need and convert other data into something useable.
I'd like to, for example:
remove all the 0 (zero) data from FKC and FLKC. Eg so only values below 0 appear on the graph…
and
turn knock, eg every knock event into a single data point, rather than the current data which simply shows knock events incrementing.

Can any logical brainiacs please help me with the formulae or describe the correct approach?
Below are examples from MegaLogViewer, one of which (Transients1) I've successfully modified to suit my btSsm logs.


RPM Change

Filter any data where the engine is experiencing a change in RPM greater than 5%, likely a transient point

abs([Field.RPM] - [RPM-4]) > ([Field.RPM] * 0.05)

Cold
Filter to identify and remove data when coolant temperature is below 160 degrees

[Field.coolantTemp] < 160

Idle Range
Filter to identify and remove data while at idle

[Field.TP] == 0 && [Field.RPM] <1000

Transients
Any data where the engine is going through transient conditions

(abs([Field.RPM] - [RPM-4]) > ([Field.RPM] * 0.10) ) || (abs([Field.TP] - [TP-4]) > 5 ) || (abs([Field.MAP] - [MAP-4]) > ([Field.MAP] * 0.15) )

Transients1
Subaru btSsm version: Any data where the engine is going through transient conditions

(abs([Field.RPM] - [RPM-4]) > ([Field.RPM] * 0.10) ) || (abs([Throttle plate] - [Throttle plate-4]) > 5 ) || (abs([Boost] - [Boost-4]) > ([Boost] * 0.15) )

[bigBADbenny]

So for example:

FKC is less than zero
Filter to identify and remove data when FKC is 0 and show all data below 0
[Field.FKC] > 0 (not working as desired)


FLKC is less than zero
Filter to identify and remove data when FLKC is 0 and show all data below 0
[Field.FLKC] > 0 (not working as desired)

Knock Event
Filter to convert knock increments into discrete data points
??? but something to do with subtracting a knock value from the next increment and making that a knock event to be used in the viewer eg in a timeline or plotted against other data.

[tremayne_xt]

Have you tried it like this

FKC is less than zero
Filter to identify and remove data when FKC is 0 and show all data below 0
[Field.FKC] = 0


FLKC is less than zero
Filter to identify and remove data when FLKC is 0 and show all data below 0
[Field.FLKC] = 0

While this would still show positive entries, that in itself would mean you have had an IAM drop so should be looked at anyway?

[bigBADbenny]

Cheers! I'll give it a go this eve and report back.
I don't have any major issues atm just looking for more clarity in the graphed logs

[bigBADbenny]

Works perfectly thanks!
Much easier to see the data I need

Any ideas for turning knock increments into data points?

Basically whenever the value changes the previous increment needs to be subtracted from the following increment leaving the actual change in value as the data....

[tremayne_xt]

Define a Custom calculated field, something like - KnockEvent=[Knocksum]-[Knocksum-1]
where [knocksum] = whatever you called it in your logs.

This should leave you with a value of 1 each time knocksum advances.

[bigBADbenny]

Love it and doing it right now
Thanks, I'm sure I'll start to "get" the way the syntax/intent works

[tremayne_xt]

Let me know if it works.

[bigBADbenny]

I made a custom field, just testing it out...
I also got the Idle Range filter working: [Field.RPM] < 1000 && [Throttle plate] < 2.5

This is what I wound up with:

[Knocks] - ([Knocks] - 1)

I think I was asking the wrong question, as this just gives a series of ones on a timeline (X axis = time, Y axis = KnockEvent)

Screen shot 2015-07-15 at 8.05.01 PM.jpg (43.17 KiB) Viewed 4468 times

You can see in this log that the Knocks increment in larger steps sometimes...
I'm trying to look at the relationship between the severity or jumps of the knock, in relation to other fields.
So rather than seeing 1 2 4 7 13 14 17 21, I want to get the magnitude of the jumps, eg: 1 2 3 6 1 3 4.
Given that the knock sum goes up to 34 then resets, this might be tricky...