How to access signal data when generators are failing?

[monoboy-hh]

Python 3.7.6
Name: asammdf
Version: 7.0.7
Linux 5.4.0-109-generic

Hi danielhrisca,
first off thanks for this great lib. I see you are rather active over here so I thought I give it shot:

Using the iter_groups and iter_channels generators I can only access two of three available channels in my mf4 (variable name 'mdf' in the following). The third one raises an exception "wrong data type "11" for vlsd channel"
for g in mdf.iter_groups(): print(f'Channels in group: {g.keys()}')
output:
Channels in group: Index(['GPSTimestamp', 'MotorSpeed'], dtype='object')
Ignoring wrong data type "11" for vlsd channel, mdf file appears to be invalid.

The mdf file was recorded with some vector logger and the vector validator claims it is legal. Despite the file being valid or not, it comes from a customer so there's not much to do about it.
However, the "missing" third channel is present in the mdf.channels_db, so I can access it through
third_channel: Channel = mdf.groups[my_group_index].channels[my_channel_index]
and can read the pointer to the Signal Data through
sig_addr = third_channel.data_block_addr

At this point I'm somewhat stuck how to finally retrieve pandas dataframes and I have the feeling it's not the right path anyway.
(Using something like
stream = mdf.select(channels=['PointCloud'])
fails with the same exception)

So my questions are:

• is there a way to work around the 'wrong data type exception'?
• or, how to 'manually' follow the pointers to the signal data to retrieve pandas df?

I can provide demo data, though about 250 MB in size.
Please find the structure attached as screenshot.

Thank you so much!
Tim

[danielhrisca]

@monoboy-hh please try the development branch code

[monoboy-hh]

Wow, thanks for the ultra-quick fix, Daniel!
version = "7.1.0.dev42" does the job

Channels in group: Index(['GPSTimestamp', 'MotorSpeed'], dtype='object') Channels in group: Index(['PointCloud'], dtype='object')

Thank you!
Best,
Tim

[YoBruce45]

hello @danielhrisca
from my side as well thank you for this very helpful library.
I am facing kind of the same issue as @monoboy-hh :

I open a mdf_file and iter through the groups with
for group in mdf_file.iter_groups()

This fails though, as one group cannot be parsed, with the exception "asammdf.blocks.utils.MdfException: wrong data type "0" for vlsd channel "ETH_Frame.DataBytes"" and I am not able to load the file at all.

Is there a way to silently ignore this kind of exception and if yes, how?
(if the fix above applies, how can i use version = "7.1.0.dev42" ?)

Thanks a lot in advance for your support and kind regards

[danielhrisca]

can you send your file? I think there is something weird about it

[danielhrisca]

for i in range(len(mdf_file.groups)):
    try:
        df = mdf_file.get_group(i)
    except Exception as e:
        print(e)

[YoBruce45]

thanks a lot, works like a charm =)