[phenixbb] AssertionError at asking for kicked maps

[Jorge Iulek]

Dear all,

     It would be nice to have kicked maps for a structure I am working 
with now. When I set then to True, I get in my log:

"
...

shake map coefficients cycles:
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
shake f_model cycles:
approx_equal eps: 1e-06
approx_equal multiplier: 10000000000.0
0.231506183818 approx_equal ERROR
0.235167389229 approx_equal ERROR

Traceback (most recent call last):
   File 
"/usr/local/phenix-1.8.4-1496/build/intel-linux-2.6-x86_64/../../phenix/phenix/command_line/refine.py", 
line 11, in <module>
     command_line.run(command_name="phenix.refine", args=sys.argv[1:])
   File 
"/usr/local/phenix-1.8.4-1496/phenix/phenix/refinement/command_line.py", 
line 177, in run
     map_manger.write_files()
   File 
"/usr/local/phenix-1.8.4-1496/phenix/phenix/refinement/driver.py", line 
1780, in write_files
     map_coeff_dataset=self.write_mtz_file(),
   File 
"/usr/local/phenix-1.8.4-1496/phenix/phenix/refinement/driver.py", line 
1766, in write_mtz_file
     pdb_hierarchy=self.refine_object.model.pdb_hierarchy())
   File 
"/usr/local/phenix-1.8.4-1496/cctbx_project/mmtbx/maps/__init__.py", 
line 515, in __init__
     pdb_hierarchy            = pdb_hierarchy)
   File 
"/usr/local/phenix-1.8.4-1496/cctbx_project/mmtbx/maps/__init__.py", 
line 439, in map_coefficients_from_fmodel
     map_type = params.map_type).map_coefficients
   File 
"/usr/local/phenix-1.8.4-1496/cctbx_project/mmtbx/maps/__init__.py", 
line 650, in __init__
     assert approx_equal(fmodel.r_work(), fmodel_dc.r_work())
AssertionError
"

     I see a previous problem like this at 
http://www.phenix-online.org/pipermail/phenixbb/2009-July/013517.html , 
but I cannot find a follow up concerning how to solve it. I am using the 
latest stable version, phenix-1.8.4-1496  , under openSuSE 12.3.

     When I set kicked maps to false, everything goes fine.
     Can this be solved? Also, a point to a former version that can 
"kick" would be enough.
     Thanks,

Jorge