I see a few issues, nothing major. 

First, your data set is missing one element.  If I strip the first-line header and then load it into Matlab, I should see a vector with 1025*1025 elements, but you are not saving the last value.  Also, the min value is negative.  So your raw data set has negative values (as well as some NaN values), which is why you are getting a negative min value in OpticStudio.  You might want to carefully check your data set to make sure it contains what you expect.

Secondly, the raw grid-phase data set has units of radians.  In OpticStudio the phase is divided by 2pi, so it is displayed in “waves.”  This explains the scaling discrepancy you observe.

Lastly, you are not displaying the full data set in OpticStudio, but instead it is being clipped by the circular system aperture.  If you open up the aperture, you will see the whole profile with the correct max value.