Merge remote branch 'y-p/fix_col_space'

* y-p/fix_col_space:
  DOC: change definition of col_space.
  BUG: df.to_html() should respect col_space argument. #1000
  TST: df.to_html() should respect col_space argument
  BUG: df.to_string() should respect self.col_space, #1000
  TST: df.to_string() should respect self.col_space

Author: wesm

doc/source/io.rst

@@ -669,7 +669,7 @@ over the string representation of the object. All arguments are optional:
 
   - ``buf`` default None, for example a StringIO object
   - ``columns`` default None, which columns to write
-  - ``col_space`` default None, number of spaces to write between columns
+  - ``col_space`` default None, minimum width of each column.
   - ``na_rep`` default ``NaN``, representation of NA value
   - ``formatters`` default None, a dictionary (by column) of functions each of
     which takes a single argument and returns a formatted string

pandas/core/format.py

@@ -30,7 +30,7 @@
     columns : sequence, optional
         the subset of columns to write; default None writes all columns
     col_space : int, optional
-        the width of each columns
+        the minimum width of each column
     header : bool, optional
         whether to print column labels, default True
     index : bool, optional
@@ -221,7 +221,7 @@ def _to_str_columns(self, force_unicode=None):
                 fmt_values = self._format_col(i)
                 cheader = str_columns[i]
 
-                max_colwidth = max(_strlen(x) for x in cheader)
+                max_colwidth = max(self.col_space or 0, *(_strlen(x) for x in cheader))
 
                 fmt_values = _make_fixed_width(fmt_values, self.justify,
                                                minimum=max_colwidth)
@@ -431,6 +431,11 @@ def write(self, s, indent=0):
         self.elements.append(' ' * indent + com.pprint_thing(s))
 
     def write_th(self, s, indent=0, tags=None):
+        if (self.fmt.col_space is not None
+            and self.fmt.col_space > 0 ):
+            tags = (tags or "" )
+            tags += 'style="min-width: %s;"' % self.fmt.col_space
+
         return self._write_cell(s, kind='th', indent=indent, tags=tags)
 
     def write_td(self, s, indent=0, tags=None):

pandas/tests/test_format.py

@@ -183,6 +183,30 @@ def test_to_string_buffer_all_unicode(self):
         # this should work
         buf.getvalue()
 
+    def test_to_string_with_col_space(self):
+        df = DataFrame(np.random.random(size=(1,3)))
+        c10=len(df.to_string(col_space=10).split("\n")[1])
+        c20=len(df.to_string(col_space=20).split("\n")[1])
+        c30=len(df.to_string(col_space=30).split("\n")[1])
+        self.assertTrue( c10 < c20 < c30 )
+
+    def test_to_html_with_col_space(self):
+        def check_with_width(df,col_space):
+            import re
+            # check that col_space affects HTML generation
+            # and be very brittle about it.
+            html = df.to_html(col_space=col_space)
+            hdrs = [x for x in html.split("\n") if re.search("<th[>\s]",x)]
+            self.assertTrue(len(hdrs) > 0 )
+            for h in hdrs:
+                self.assertTrue("min-width" in h )
+                self.assertTrue(str(col_space) in h )
+
+        df = DataFrame(np.random.random(size=(1,3)))
+
+        check_with_width(df,30)
+        check_with_width(df,50)
+
     def test_to_html_unicode(self):
         # it works!
         df = DataFrame({u'\u03c3' : np.arange(10.)})