diff --git a/processing/pre/geom_clean.py b/processing/pre/geom_clean.py
index 7ee87fe0a..50f3657b2 100755
--- a/processing/pre/geom_clean.py
+++ b/processing/pre/geom_clean.py
@@ -5,9 +5,6 @@ import sys
 from io import StringIO
 from optparse import OptionParser
 
-from scipy import ndimage
-import numpy as np
-
 import damask
 
 
diff --git a/processing/pre/geom_mirror.py b/processing/pre/geom_mirror.py
index 3f9755818..77ec1f4d7 100755
--- a/processing/pre/geom_mirror.py
+++ b/processing/pre/geom_mirror.py
@@ -5,8 +5,6 @@ import sys
 from io import StringIO
 from optparse import OptionParser
 
-import numpy as np
-
 import damask
 
 
diff --git a/python/damask/geom.py b/python/damask/geom.py
index 6fb9af5a5..6dfcda013 100644
--- a/python/damask/geom.py
+++ b/python/damask/geom.py
@@ -311,8 +311,8 @@ class Geom():
         """
         header = self.get_header()
         grid   = self.get_grid()
-        format_string = '%{}i'.format(1+int(np.floor(np.log10(np.nanmax(self.microstructure))))) if self.microstructure.dtype == int \
-                   else '%g'
+        format_string = '%g' if self.microstructure in np.sctypes['float'] else \
+                        '%{}i'.format(1+int(np.floor(np.log10(np.nanmax(self.microstructure)))))
         np.savetxt(fname,
                    self.microstructure.reshape([grid[0],np.prod(grid[1:])],order='F').T,
                    header='\n'.join(header), fmt=format_string, comments='')
@@ -390,12 +390,14 @@ class Geom():
     def mirror(self,directions,reflect=False):
         """
         Mirror microstructure along given directions.
+
         Parameters
         ----------
         directions : iterable containing str
             direction(s) along which the microstructure is mirrored. Valid entries are 'x', 'y', 'z'.
         reflect : bool, optional
             reflect (include) outermost layers.
+
         """
         valid = {'x','y','z'}
         if not all(isinstance(d, str) for d in directions):
@@ -420,10 +422,12 @@ class Geom():
     def clean(self,stencil=3):
         """
         Smooth microstructure by selecting most frequent index within given stencil at each location.
+
         Parameters
         ----------
         stencil : int, optional
             size of smoothing stencil.
+
         """
         def mostFrequent(arr):
             unique, inverse = np.unique(arr, return_inverse=True)