diff --git a/pandas/core/format.py b/pandas/core/format.py
index ae0d95b1c3074..6ba7bfcc8f16c 100644
--- a/pandas/core/format.py
+++ b/pandas/core/format.py
@@ -62,6 +62,7 @@
-------
formatted : string (or unicode, depending on data and options)"""
+
class CategoricalFormatter(object):
def __init__(self, categorical, buf=None, length=True,
na_rep='NaN', name=False, footer=True):
@@ -140,7 +141,7 @@ def __init__(self, series, buf=None, header=True, length=True,
if float_format is None:
float_format = get_option("display.float_format")
self.float_format = float_format
- self.dtype = dtype
+ self.dtype = dtype
def _get_footer(self):
footer = u('')
@@ -163,7 +164,7 @@ def _get_footer(self):
footer += 'Length: %d' % len(self.series)
if self.dtype:
- if getattr(self.series.dtype,'name',None):
+ if getattr(self.series.dtype, 'name', None):
if footer:
footer += ', '
footer += 'dtype: %s' % com.pprint_thing(self.series.dtype.name)
@@ -213,6 +214,7 @@ def to_string(self):
return compat.text_type(u('\n').join(result))
+
def _strlen_func():
if compat.PY3: # pragma: no cover
_strlen = len
@@ -304,32 +306,31 @@ def _to_str_columns(self):
for i, c in enumerate(self.columns):
if self.header:
- fmt_values = self._format_col(i)
cheader = str_columns[i]
-
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)
+ fmt_values = self._format_col(i, justify=self.justify,
+ minimum=max_colwidth)
max_len = max(np.max([_strlen(x) for x in fmt_values]),
max_colwidth)
if self.justify == 'left':
cheader = [x.ljust(max_len) for x in cheader]
- else:
+ elif self.justify == 'right':
cheader = [x.rjust(max_len) for x in cheader]
+ elif self.justify == 'center':
+ cheader = [x.center(max_len) for x in cheader]
+ else:
+ cheader = [x.strip() for x in cheader]
stringified.append(cheader + fmt_values)
else:
- stringified = [_make_fixed_width(self._format_col(i),
- self.justify)
+ stringified = [self._format_col(i, justify=self.justify)
for i, c in enumerate(self.columns)]
strcols = stringified
if self.index:
strcols.insert(0, str_index)
-
return strcols
def to_string(self, force_unicode=None):
@@ -450,12 +451,14 @@ def write(buf, frame, column_format, strcols):
raise TypeError('buf is not a file name and it has no write '
'method')
- def _format_col(self, i):
+ def _format_col(self, i, justify='right', minimum=None):
formatter = self._get_formatter(i)
return format_array(self.frame.icol(i).get_values(), formatter,
float_format=self.float_format,
na_rep=self.na_rep,
- space=self.col_space)
+ space=self.col_space,
+ justify=justify,
+ minimum=minimum)
def to_html(self, classes=None):
"""
@@ -735,7 +738,7 @@ def _write_body(self, indent):
fmt_values = {}
for i in range(len(self.columns)):
- fmt_values[i] = self.fmt._format_col(i)
+ fmt_values[i] = self.fmt._format_col(i, justify=None)
# write values
if self.fmt.index:
@@ -1485,7 +1488,7 @@ def get_formatted_cells(self):
def format_array(values, formatter, float_format=None, na_rep='NaN',
- digits=None, space=None, justify='right'):
+ digits=None, space=None, justify='right', minimum=None):
if com.is_float_dtype(values.dtype):
fmt_klass = FloatArrayFormatter
elif com.is_integer_dtype(values.dtype):
@@ -1509,7 +1512,7 @@ def format_array(values, formatter, float_format=None, na_rep='NaN',
fmt_obj = fmt_klass(values, digits, na_rep=na_rep,
float_format=float_format,
formatter=formatter, space=space,
- justify=justify)
+ justify=justify, minimum=minimum)
return fmt_obj.get_result()
@@ -1517,7 +1520,7 @@ def format_array(values, formatter, float_format=None, na_rep='NaN',
class GenericArrayFormatter(object):
def __init__(self, values, digits=7, formatter=None, na_rep='NaN',
- space=12, float_format=None, justify='right'):
+ space=12, float_format=None, justify='right', minimum=None):
self.values = values
self.digits = digits
self.na_rep = na_rep
@@ -1525,10 +1528,11 @@ def __init__(self, values, digits=7, formatter=None, na_rep='NaN',
self.formatter = formatter
self.float_format = float_format
self.justify = justify
+ self.minimum = minimum
def get_result(self):
fmt_values = self._format_strings()
- return _make_fixed_width(fmt_values, self.justify)
+ return _make_fixed_width(fmt_values, self.justify, self.minimum)
def _format_strings(self):
if self.float_format is None:
@@ -1584,19 +1588,19 @@ def __init__(self, *args, **kwargs):
def _format_with(self, fmt_str):
def _val(x, threshold):
if notnull(x):
- if threshold is None or abs(x) > get_option("display.chop_threshold"):
+ if threshold is None or abs(x) > get_option("display.chop_threshold"):
return fmt_str % x
else:
- if fmt_str.endswith("e"): # engineering format
- return "0"
+ if fmt_str.endswith("e"): # engineering format
+ return "0"
else:
- return fmt_str % 0
+ return fmt_str % 0
else:
return self.na_rep
threshold = get_option("display.chop_threshold")
- fmt_values = [ _val(x, threshold) for x in self.values]
+ fmt_values = [_val(x, threshold) for x in self.values]
return _trim_zeros(fmt_values, self.na_rep)
def get_result(self):
@@ -1627,7 +1631,7 @@ def get_result(self):
fmt_str = '%% .%de' % (self.digits - 1)
fmt_values = self._format_with(fmt_str)
- return _make_fixed_width(fmt_values, self.justify)
+ return _make_fixed_width(fmt_values, self.justify, self.minimum)
class IntArrayFormatter(GenericArrayFormatter):
@@ -1640,7 +1644,7 @@ def get_result(self):
fmt_values = [formatter(x) for x in self.values]
- return _make_fixed_width(fmt_values, self.justify)
+ return _make_fixed_width(fmt_values, self.justify, self.minimum)
class Datetime64Formatter(GenericArrayFormatter):
@@ -1652,7 +1656,8 @@ def get_result(self):
formatter = _format_datetime64
fmt_values = [formatter(x) for x in self.values]
- return _make_fixed_width(fmt_values, self.justify)
+ return _make_fixed_width(fmt_values, self.justify, self.minimum)
+
def _format_datetime64(x, tz=None):
if isnull(x):
@@ -1672,7 +1677,8 @@ def get_result(self):
formatter = _format_timedelta64
fmt_values = [formatter(x) for x in self.values]
- return _make_fixed_width(fmt_values, self.justify)
+ return _make_fixed_width(fmt_values, self.justify, self.minimum)
+
def _format_timedelta64(x):
if isnull(x):
@@ -1680,6 +1686,7 @@ def _format_timedelta64(x):
return lib.repr_timedelta64(x)
+
def _make_fixed_width(strings, justify='right', minimum=None):
if len(strings) == 0:
return strings
@@ -1697,8 +1704,12 @@ def _make_fixed_width(strings, justify='right', minimum=None):
if justify == 'left':
justfunc = lambda self, x: self.ljust(x)
- else:
+ elif justify == 'right':
justfunc = lambda self, x: self.rjust(x)
+ elif justify == 'center':
+ justfunc = lambda self, x: self.center(x)
+ else:
+ justfunc = lambda self, _: self.strip()
def just(x):
eff_len = max_len
diff --git a/pandas/tests/test_format.py b/pandas/tests/test_format.py
index d9bf8adb71298..f3b09b227cc10 100644
--- a/pandas/tests/test_format.py
+++ b/pandas/tests/test_format.py
@@ -399,13 +399,13 @@ def test_to_html_escaped(self):
| str<ing1 & |
- <type 'str'> |
- <type 'str'> |
+ <type 'str'> |
+ <type 'str'> |
| stri>ng2 & |
- <type 'str'> |
- <type 'str'> |
+ <type 'str'> |
+ <type 'str'> |
"""
@@ -431,13 +431,13 @@ def test_to_html_escape_disabled(self):
str
- | bold |
- bold |
+ bold |
+ bold |
|
| stri>ng2 & |
- bold |
- bold |
+ bold |
+ bold |
"""
@@ -471,26 +471,26 @@ def test_to_html_multiindex_sparsify_false_multi_sparse(self):
| 0 |
0 |
- 0 |
- 1 |
+ 0 |
+ 1 |
| 0 |
1 |
- 2 |
- 3 |
+ 2 |
+ 3 |
| 1 |
0 |
- 4 |
- 5 |
+ 4 |
+ 5 |
| 1 |
1 |
- 6 |
- 7 |
+ 6 |
+ 7 |
"""
@@ -526,26 +526,26 @@ def test_to_html_multiindex_sparsify_false_multi_sparse(self):
| 0 |
0 |
- 0 |
- 1 |
+ 0 |
+ 1 |
| 0 |
1 |
- 2 |
- 3 |
+ 2 |
+ 3 |
| 1 |
0 |
- 4 |
- 5 |
+ 4 |
+ 5 |
| 1 |
1 |
- 6 |
- 7 |
+ 6 |
+ 7 |
"""
@@ -577,24 +577,24 @@ def test_to_html_multiindex_sparsify(self):
| 0 |
0 |
- 0 |
- 1 |
+ 0 |
+ 1 |
| 1 |
- 2 |
- 3 |
+ 2 |
+ 3 |
| 1 |
0 |
- 4 |
- 5 |
+ 4 |
+ 5 |
| 1 |
- 6 |
- 7 |
+ 6 |
+ 7 |
"""
@@ -630,24 +630,24 @@ def test_to_html_multiindex_sparsify(self):
| 0 |
0 |
- 0 |
- 1 |
+ 0 |
+ 1 |
| 1 |
- 2 |
- 3 |
+ 2 |
+ 3 |
| 1 |
0 |
- 4 |
- 5 |
+ 4 |
+ 5 |
| 1 |
- 6 |
- 7 |
+ 6 |
+ 7 |
"""
@@ -671,23 +671,23 @@ def test_to_html_index_formatter(self):
| a |
- 0 |
- 1 |
+ 0 |
+ 1 |
| b |
- 2 |
- 3 |
+ 2 |
+ 3 |
| c |
- 4 |
- 5 |
+ 4 |
+ 5 |
| d |
- 6 |
- 7 |
+ 6 |
+ 7 |
"""
@@ -1159,7 +1159,7 @@ def test_to_string_left_justify_cols(self):
df_s = df.to_string(justify='left')
expected = (' x \n'
'0 3234.000\n'
- '1 0.253')
+ '1 0.253 ')
assert(df_s == expected)
def test_to_string_format_na(self):
@@ -1274,17 +1274,17 @@ def test_to_html_multiindex(self):
' \n'
' \n'
' | 0 | \n'
- ' a | \n'
- ' b | \n'
- ' c | \n'
- ' d | \n'
+ ' a | \n'
+ ' b | \n'
+ ' c | \n'
+ ' d | \n'
'
\n'
' \n'
' | 1 | \n'
- ' e | \n'
- ' f | \n'
- ' g | \n'
- ' h | \n'
+ ' e | \n'
+ ' f | \n'
+ ' g | \n'
+ ' h | \n'
'
\n'
' \n'
'')
@@ -1316,17 +1316,17 @@ def test_to_html_multiindex(self):
' \n'
' \n'
' | 0 | \n'
- ' a | \n'
- ' b | \n'
- ' c | \n'
- ' d | \n'
+ ' a | \n'
+ ' b | \n'
+ ' c | \n'
+ ' d | \n'
'
\n'
' \n'
' | 1 | \n'
- ' e | \n'
- ' f | \n'
- ' g | \n'
- ' h | \n'
+ ' e | \n'
+ ' f | \n'
+ ' g | \n'
+ ' h | \n'
'
\n'
' \n'
'')
@@ -1351,21 +1351,21 @@ def test_to_html_justify(self):
' \n'
' \n'
' | 0 | \n'
- ' 6 | \n'
- ' 1 | \n'
- ' 223442 | \n'
+ ' 6 | \n'
+ ' 1 | \n'
+ ' 223442 | \n'
'
\n'
' \n'
' | 1 | \n'
- ' 30000 | \n'
- ' 2 | \n'
- ' 0 | \n'
+ ' 30000 | \n'
+ ' 2 | \n'
+ ' 0 | \n'
'
\n'
' \n'
' | 2 | \n'
- ' 2 | \n'
- ' 70000 | \n'
- ' 1 | \n'
+ ' 2 | \n'
+ ' 70000 | \n'
+ ' 1 | \n'
'
\n'
' \n'
'')
@@ -1385,21 +1385,21 @@ def test_to_html_justify(self):
' \n'
' \n'
' | 0 | \n'
- ' 6 | \n'
- ' 1 | \n'
- ' 223442 | \n'
+ ' 6 | \n'
+ ' 1 | \n'
+ ' 223442 | \n'
'
\n'
' \n'
' | 1 | \n'
- ' 30000 | \n'
- ' 2 | \n'
- ' 0 | \n'
+ ' 30000 | \n'
+ ' 2 | \n'
+ ' 0 | \n'
'
\n'
' \n'
' | 2 | \n'
- ' 2 | \n'
- ' 70000 | \n'
- ' 1 | \n'
+ ' 2 | \n'
+ ' 70000 | \n'
+ ' 1 | \n'
'
\n'
' \n'
'')
@@ -1934,6 +1934,69 @@ def test_misc(self):
obj = fmt.FloatArrayFormatter(np.array([], dtype=np.float64))
result = obj.get_result()
+
+class TestDataFrameJustification(unittest.TestCase):
+ def setUp(self):
+ self.df_int = pd.DataFrame(np.arange(3).reshape(1, 3),
+ columns=['foo', 'bar', 'baz'],
+ dtype=int)
+ self.df_float = pd.DataFrame(np.linspace(0.3, 1.2, 3).reshape(1, 3),
+ columns=['a', 'b', 'c'],
+ dtype=float)
+ self.df_string = pd.DataFrame([
+ ['test', 'something long', 'something even longer'],
+ ['foo', 'bar', 'baz'],
+ ['small', 'text', 'samples']
+ ], columns=['a', 'b', 'c'])
+
+ def test_left_justification(self):
+ expected = ' foo bar baz\n0 0 1 2 '
+ self.assertEqual(expected, self.df_int.to_string(justify='left'),
+ 'Left int justification failed')
+ expected = ' a b c \n0 0.3 0.75 1.2'
+ self.assertEqual(expected, self.df_float.to_string(justify='left'),
+ 'Left float justification failed')
+ expected = '''
+ a b c
+0 test something long something even longer
+1 foo bar baz
+2 small text samples
+'''.strip('\r\n')
+ self.assertEqual(expected, self.df_string.to_string(justify='left'),
+ 'Left string justification failed')
+
+ def test_right_justification(self):
+ expected = ' foo bar baz\n0 0 1 2'
+ self.assertEqual(expected, self.df_int.to_string(justify='right'),
+ 'Right int justification failed')
+ expected = ' a b c\n0 0.3 0.75 1.2'
+ self.assertEqual(expected, self.df_float.to_string(justify='right'),
+ 'Right float justification failed')
+ expected = '''
+ a b c
+0 test something long something even longer
+1 foo bar baz
+2 small text samples
+'''.strip('\r\n')
+ self.assertEqual(expected, self.df_string.to_string(justify='right'),
+ 'Right string justification failed')
+
+ def test_center_justification(self):
+ expected = ' foo bar baz\n0 0 1 2 '
+ self.assertEqual(expected, self.df_int.to_string(justify='center'),
+ 'Center int justification failed')
+ expected = ' a b c \n0 0.3 0.75 1.2'
+ self.assertEqual(expected, self.df_float.to_string(justify='center'),
+ 'Center float justification failed')
+ expected = '''
+ a b c
+0 test something long something even longer
+1 foo bar baz
+2 small text samples
+'''.strip('\r\n')
+ self.assertEqual(expected, self.df_string.to_string(justify='center'),
+ 'Center string justification failed')
+
if __name__ == '__main__':
import nose
nose.runmodule(argv=[__file__, '-vvs', '-x', '--pdb', '--pdb-failure'],