New issue 1295: [GDF] writing and loading dataset yields different result https://bitbucket.org/yt_analysis/yt/issues/1295/gdf-writing-and-loading-dat... Kacper Kowalik: As reported by Andrea Negri on the [yt-user ml](http://lists.spacepope.org/pipermail/yt-users-spacepope.org/2016-November/00...) following script: ``` #!/usr/bin/env python # -*- coding: utf-8 -*- import numpy as np import yt def main(): dims = np.array([2, 2, 2]) grid_data = [ dict(left_edge = [0.0, 0.0, 0.0], right_edge = [1.0, 1.0, 1.0], level = 0, dimensions = dims), dict(left_edge = [0.25, 0.25, 0.25], right_edge = [0.75, 0.75, 0.75], level = 1, dimensions = dims), ] bbox = np.array([[0, 1], [0, 1], [0, 1]]) ds = yt.load_amr_grids(grid_data, dims, bbox=bbox, periodicity=(0,0,0), \ length_unit=1., refine_by=2) #make a nice slice slc = yt.SlicePlot(ds, "z", "grid_level") slc.set_log("grid_level", False) slc.save('original_slice.png') yt.utilities.grid_data_format.writer.write_to_gdf(ds, 'test.h5', clobber=True) da = yt.load('test.h5') slc = yt.SlicePlot(da, "z", "grid_level") slc.set_log("grid_level", False) slc.save('test.png') if __name__ == '__main__': main() ``` yields different results ![original_slice.png] (https://bitbucket.org/repo/BXbAb/images/355574044-original_slice.png)  for the same dataset that was exported and read as a GDF file. Responsible: xarthisius