Unknown Datatype with dtype_a 0x08 and dtype 8

[FabioCurto]

Hi! Very good parser, but I am facing a problem trying to parse a log file from a MoTeC M130 ecu of a motorbike that outputs the ld and ldx file. Almost every signal falls into the expected cases with the datatype as in the screenshot:

The problem is that I need the GPS.Latitude and GPS.Longitude signals which instead does not fall into no case since it has dtype_a = 8 and dtype = 8.

Do you know which datatype is needed or how to interpret it? Or at least how to find out from the binary file?

Also I noticed that the parsing of the header is not perfectly done for this signal since its unit is empty and the short name is "deg" and I'm thinking the problem could be there

Thanks!

[gotzl]

Hi,
the issue with the unit could be related to #18, I forgot to look into this ... ^^
The only way I know to find out what dtype_a/dtype combination represents is to look into the binary file where the data is located and take a guess.. If you can provide me a sample of your .ld file I can give it a try.

[FabioCurto]

Surely I can provide you the sample .ld file. It is quite heavy (~80 Mb) and github does not allow any .ld file to upload here. How can I share it with you?

Note that If I have to guess what to expect the Latitude and Longitude to be, is something like (45.08528382440717, 10.312579068846695) that are the coordinates where I recorded the sample file!

[gotzl]

What you could do is modify the code such dtype = np.uint8 is used, and then post some values as hex here. This could be enough for me to figure out how the GPS coords are encoded.
S.t. like

import binascii
from ldparser import ldData
l = ldData.fromfile('/tmp/test.ld')
print(binascii.hexlify(l[0].data[:64]))

[FabioCurto]

Hi! I had to pick up the last 64 elements since at the beginning the GPS was not fixed and was retrieveng all zeros:
Latitude:
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

Longitude:
0000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c400000000000006440000000000000424000000000000050400000000000e06240000000000080524000000000002061400000000000805b400000000000001c40000000000000644000000000000042400000000000005040000000000080694000000000008047400000000000804c40

[gotzl]

Hmm... don't recognize s.t. at first glance. Could you comment https://github.com/gotzl/ldparser/blob/master/ldparser.py#L397 just to be sure that this line didn't alter the data?

[FabioCurto]

It was somehow modifying the data, Now I have as a result:

Latitude:
a6f58a4640f9a9e0a6f58a4640f9a9e0a6f58a4640f9a9e0a6f58a464046a34cb4f58a464046a34cb4f58a464046a34cb4f58a464046a34cb4f58a4640295c8f

Longitude:
6e07a02440974a896e07a02440974a896e07a02440974a896e07a02440974a896e07a02440974a896e07a02440974a896e07a02440974a896e07a02440cc2f3

Just to be complete, these are its parameters extracted for both the signals:
freq = 20,
shift = 0,
mul = 1,
scale = 1,
dec = 0

[FabioCurto]

UPDATE:

Looking through the Latitude, I found this two repeated sequences "46a34cb4f58a4640" and "f9a9e0a6f58a4640" that could resemble respectively 45.085623299999995° and 45.0856217° if interpreted as double floating point and with little endianess. Then trying to find some similarities in the longitude, I found "974a896e07a02440" that is converted in 10.3125567°. Could this be the right path to follow?

[gotzl]

Hi, yes, this sounds very good! This would mean dtype = np.dtype('<f8')...

[FabioCurto]

Happy to say that it worked (avoiding some lost signal to 0 that i will filter on)!

I did not use np.dtype('<f8') but used np.dtype('<d') and it works perfectly. Thanks for the help!