Shape table moments - how are they calculated?

[jling-NM]

I am trying to verify statistical moments in mindboggle output “sulcus_shapes.csv”.

To do so, select a single structure from the file and focus on area and travel depth columns here:
|name|ID|area|travel depth: median|travel depth: mean|
|collateral sulcus|23|2349.66863596863|5.9590169098|6.48712380995582|
Note area and median values.

Now, open file “vertices.csv” at same location. Select all rows where sulcus ID = 23
|label ID|sulcus ID|area|travel depth|
|1016|23|0.84241010249|7.7849208317|
…
etc.

When i calculate the sum of the area (2349.66863596863) it matches “sulcus_shapes.csv”. When i calculate moments like mean(7.1330239621) and median(7.34413325939166), the values do not match “sulcus_shapes.csv”.
Can anyone explain the error in my method or where these statistical moments originate?

thanks

[PeerHerholz]

Ahoi hoi @jling-NM,

thank you very much for your post.
Mindboggle's method to generate the summary tables can be found here:
https://github.com/nipy/mindboggle/blob/master/mindboggle/mio/tables.py .

As for your example: could you maybe provide further details like the function you are using for computing these values (please have in mind that a lot of parts can make a difference: the precise function, the version, etc. ) and how much they deviate? Would it be possible to include example data with the code you used so that we can try to replicate what's going on?

Best, Peer

[jling-NM]

Thank you. These table values appear to be area-weighted when I look at https://github.com/nipy/mindboggle/blob/master/mindboggle/mio/tables.py