An error was encountered while fitting DKI

[406354348]

Traceback (most recent call last):
File "/home/jht/anaconda3/envs/gnn/lib/python3.7/multiprocessing/queues.py", line 242, in _feed
send_bytes(obj)
File "/home/jht/anaconda3/envs/gnn/lib/python3.7/multiprocessing/connection.py", line 200, in send_bytes
self._send_bytes(m[offset:offset + size])
File "/home/jht/anaconda3/envs/gnn/lib/python3.7/multiprocessing/connection.py", line 393, in _send_bytes
header = struct.pack("!i", n)
struct.error: 'i' format requires -2147483648 <= number <= 2147483647

[kodiweera]

Can you please try fitting dki on a single data set? Do you have sufficient RAM?

[406354348]

Can you please try fitting dki on a single data set? Do you have sufficient RAM?
I just use one data for fitting and have enough RAM, while my DMRI data is HCP, and the data dimension is (145,174,145,288)

[kodiweera]

How many shells your dwi data have? Did you correclty mentioned number of b0 and b-values? How did your DTI fitting go?

[406354348]

你的 dwi 数据有多少个 shell？您是否正确提到了 b0 和 b 值的数量？你的 DTI 试衣效果如何？

My DWI data has three shells, each with 90 directions, b1000, b2000 and b3000, and there are 18 in b0. The code I use is python -m difit '/home/jht/gnn_pnt/data/102311' '/home/jht/gnn_pnt/data/102311/out' '/home/jht/gnn_pnt/data/work' --models dki --dki_b_values 1000 2000 3000 --dki_b0_images 18 --mem 9 --nprocs 1. DTI useless attempt. It is also possible that mem is set to 9, which is too large.

[kodiweera]

Please give it a try with fewer b0 images. increase --nprocs to 2. If you have more RAM, you can increase the ram. How big is the dataset?

[kodiweera]

You can try the docker/ singularity image to single out any system related issues. If you do a longitudinal study, container approach is the best.

[406354348]

Thank you for your answer, my data size is 1.15g, I'll try your comments.