MT-1 optic flow typo

Hi all, I found a typo in the MT-1 data set. According to both the paper and the GetVelField script, the columns in the mtdata.opticflow matrix should represent (in order) horizontal and vertical translation, expansion, rotation, and horizontal and vertical shear. However, I found that the last four columns are actually switched; the 3rd and 4th columns actually represent shear in the visual stimulus, while the 5th and 6th columns contain expansion and rotation, respectively. Just a heads up.

- Jio