no sensible output in recon_ismrmrd_dataset.py

[zwep]

I just tried your script.. I just got noise that was based off of data provided by the philips_to_ismrmrd tool...

Do you have a dataset (.h5) that works for you and that you can share? That would be great.

Also.. I don't understand why you are performing this statement:

       xline = htransf.transform_kspace_to_image(acq.data, [1])
        x0 = (eNx - rNx) / 2
        x1 = (eNx - rNx) / 2 + rNx
        xline = xline[:, int(x0):int(x1)]
        acq.resize(rNx, acq.active_channels, acq.trajectory_dimensions)
        acq.center_sample = rNx // 2
        # need to use the [:] notation here to fill the data
        acq.data[:] = htransf.transform_image_to_kspace(xline, [1])

Why are you transforming each line to image space.. and then back to kspace.. Is this really the most efficient way to correct for oversampling?

Cant you just specify the target image space when you've collected all the kspace data, and give that to transform_image_to_kspace?