示例#1
0
    def test_different_table_rows(self):
        """
        Test tables that are otherwise identical but one has more rows than the
        other.
        """

        ca1 = Column('A', format='L', array=[True, False])
        cb1 = Column('B', format='L', array=[True, False])
        ca2 = Column('A', format='L', array=[True, False, True])
        cb2 = Column('B', format='L', array=[True, False, True])

        ta = BinTableHDU.from_columns([ca1, cb1])
        tb = BinTableHDU.from_columns([ca2, cb2])

        diff = TableDataDiff(ta.data, tb.data)

        assert not diff.identical
        assert diff.diff_column_count == ()
        assert len(diff.common_columns) == 2
        assert diff.diff_rows == (2, 3)
        assert diff.diff_values == []

        report = diff.report()

        assert 'Table rows differ' in report
        assert 'a: 2' in report
        assert 'b: 3' in report
        assert 'No further data comparison performed.'
示例#2
0
    def test_different_table_field_counts(self):
        """
        Test tables with some common columns, but different number of columns
        overall.
        """

        ca = Column('A', format='L', array=[True, False])
        cb = Column('B', format='L', array=[True, False])
        cc = Column('C', format='L', array=[True, False])

        ta = BinTableHDU.from_columns([cb])
        tb = BinTableHDU.from_columns([ca, cb, cc])

        diff = TableDataDiff(ta.data, tb.data)

        assert not diff.identical
        assert diff.diff_column_count == (1, 3)
        assert len(diff.common_columns) == 1
        assert diff.common_column_names == {'b'}
        assert diff.diff_column_names == ([], ['A', 'C'])
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0

        report = diff.report()
        assert ' Tables have different number of columns:' in report
        assert '  a: 1\n  b: 3' in report
示例#3
0
    def test_partially_identical_files3(self):
        """
        Test files that have some identical HDUs but a different extension
        name.
        """

        phdu = PrimaryHDU()
        ehdu = ImageHDU(name='FOO')
        hdula = HDUList([phdu, ehdu])
        ehdu = BinTableHDU(name='BAR')
        ehdu.header['EXTVER'] = 2
        ehdu.header['EXTLEVEL'] = 3
        hdulb = HDUList([phdu, ehdu])
        diff = FITSDiff(hdula, hdulb)
        assert not diff.identical

        assert diff.diff_hdus[0][0] == 1

        hdu_diff = diff.diff_hdus[0][1]
        assert hdu_diff.diff_extension_types == ('IMAGE', 'BINTABLE')
        assert hdu_diff.diff_extnames == ('FOO', 'BAR')
        assert hdu_diff.diff_extvers == (1, 2)
        assert hdu_diff.diff_extlevels == (1, 3)

        report = diff.report()
        assert 'Extension types differ' in report
        assert 'a: IMAGE\n    b: BINTABLE' in report
        assert 'Extension names differ' in report
        assert 'a: FOO\n    b: BAR' in report
        assert 'Extension versions differ' in report
        assert 'a: 1\n    b: 2' in report
        assert 'Extension levels differ' in report
        assert 'a: 1\n    b: 2' in report
示例#4
0
    def test_different_table_field_names(self):
        ca = Column('A', format='L', array=[True, False])
        cb = Column('B', format='L', array=[True, False])
        cc = Column('C', format='L', array=[True, False])

        ta = BinTableHDU.from_columns([ca, cb])
        tb = BinTableHDU.from_columns([ca, cc])

        diff = TableDataDiff(ta.data, tb.data)

        assert not diff.identical
        assert len(diff.common_columns) == 1
        assert diff.common_column_names == {'a'}
        assert diff.diff_column_names == (['B'], ['C'])
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0

        report = diff.report()
        assert 'Extra column B of format L in a' in report
        assert 'Extra column C of format L in b' in report
示例#5
0
    def test_different_table_field_names(self):
        ca = Column('A', format='L', array=[True, False])
        cb = Column('B', format='L', array=[True, False])
        cc = Column('C', format='L', array=[True, False])

        ta = BinTableHDU.from_columns([ca, cb])
        tb = BinTableHDU.from_columns([ca, cc])

        diff = TableDataDiff(ta.data, tb.data)

        assert not diff.identical
        assert len(diff.common_columns) == 1
        assert diff.common_column_names == {'a'}
        assert diff.diff_column_names == (['B'], ['C'])
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0

        report = diff.report()
        assert 'Extra column B of format L in a' in report
        assert 'Extra column C of format L in b' in report
示例#6
0
    def test_ignore_hdus(self):
        a = np.arange(100).reshape(10, 10)
        b = a.copy()
        ha = Header([('A', 1), ('B', 2), ('C', 3)])
        xa = np.array([(1.0, 1), (3.0, 4)], dtype=[('x', float), ('y', int)])
        xb = np.array([(1.0, 2), (3.0, 5)], dtype=[('x', float), ('y', int)])
        phdu = PrimaryHDU(header=ha)
        ihdua = ImageHDU(data=a, name='SCI')
        ihdub = ImageHDU(data=b, name='SCI')
        bhdu1 = BinTableHDU(data=xa, name='ASDF')
        bhdu2 = BinTableHDU(data=xb, name='ASDF')
        hdula = HDUList([phdu, ihdua, bhdu1])
        hdulb = HDUList([phdu, ihdub, bhdu2])

        # ASDF extension should be different
        diff = FITSDiff(hdula, hdulb)
        assert not diff.identical
        assert diff.diff_hdus[0][0] == 2

        # ASDF extension should be ignored
        diff = FITSDiff(hdula, hdulb, ignore_hdus=['ASDF'])
        assert diff.identical, diff.report()

        diff = FITSDiff(hdula, hdulb, ignore_hdus=['ASD*'])
        assert diff.identical, diff.report()

        # SCI extension should be different
        hdulb['SCI'].data += 1
        diff = FITSDiff(hdula, hdulb, ignore_hdus=['ASDF'])
        assert not diff.identical

        # SCI and ASDF extensions should be ignored
        diff = FITSDiff(hdula, hdulb, ignore_hdus=['SCI', 'ASDF'])
        assert diff.identical, diff.report()

        # All EXTVER of SCI should be ignored
        ihduc = ImageHDU(data=a, name='SCI', ver=2)
        hdulb.append(ihduc)
        diff = FITSDiff(hdula, hdulb, ignore_hdus=['SCI', 'ASDF'])
        assert not any(diff.diff_hdus), diff.report()
        assert any(diff.diff_hdu_count), diff.report()
示例#7
0
    def test_ignore_table_fields(self):
        c1 = Column('A', format='L', array=[True, False])
        c2 = Column('B', format='X', array=[[0], [1]])
        c3 = Column('C', format='4I', dim='(2, 2)',
                    array=[[0, 1, 2, 3], [4, 5, 6, 7]])

        c4 = Column('B', format='X', array=[[1], [0]])
        c5 = Column('C', format='4I', dim='(2, 2)',
                    array=[[1, 2, 3, 4], [5, 6, 7, 8]])

        ta = BinTableHDU.from_columns([c1, c2, c3])
        tb = BinTableHDU.from_columns([c1, c4, c5])

        diff = TableDataDiff(ta.data, tb.data, ignore_fields=['B', 'C'])
        assert diff.identical

        # The only common column should be c1
        assert len(diff.common_columns) == 1
        assert diff.common_column_names == {'a'}
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0
示例#8
0
    def test_ignore_table_fields(self):
        c1 = Column('A', format='L', array=[True, False])
        c2 = Column('B', format='X', array=[[0], [1]])
        c3 = Column('C', format='4I', dim='(2, 2)',
                    array=[[0, 1, 2, 3], [4, 5, 6, 7]])

        c4 = Column('B', format='X', array=[[1], [0]])
        c5 = Column('C', format='4I', dim='(2, 2)',
                    array=[[1, 2, 3, 4], [5, 6, 7, 8]])

        ta = BinTableHDU.from_columns([c1, c2, c3])
        tb = BinTableHDU.from_columns([c1, c4, c5])

        diff = TableDataDiff(ta.data, tb.data, ignore_fields=['B', 'C'])
        assert diff.identical

        # The only common column should be c1
        assert len(diff.common_columns) == 1
        assert diff.common_column_names == {'a'}
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0
示例#9
0
    def test_identical_tables(self):
        c1 = Column('A', format='L', array=[True, False])
        c2 = Column('B', format='X', array=[[0], [1]])
        c3 = Column('C', format='4I', dim='(2, 2)',
                    array=[[0, 1, 2, 3], [4, 5, 6, 7]])
        c4 = Column('D', format='J', bscale=2.0, array=[0, 1])
        c5 = Column('E', format='A3', array=['abc', 'def'])
        c6 = Column('F', format='E', unit='m', array=[0.0, 1.0])
        c7 = Column('G', format='D', bzero=-0.1, array=[0.0, 1.0])
        c8 = Column('H', format='C', array=[0.0+1.0j, 2.0+3.0j])
        c9 = Column('I', format='M', array=[4.0+5.0j, 6.0+7.0j])
        c10 = Column('J', format='PI(2)', array=[[0, 1], [2, 3]])

        columns = [c1, c2, c3, c4, c5, c6, c7, c8, c9, c10]

        ta = BinTableHDU.from_columns(columns)
        tb = BinTableHDU.from_columns([c.copy() for c in columns])

        diff = TableDataDiff(ta.data, tb.data)
        assert diff.identical
        assert len(diff.common_columns) == 10
        assert diff.common_column_names == set('abcdefghij')
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0
示例#10
0
    def test_identical_tables(self):
        c1 = Column('A', format='L', array=[True, False])
        c2 = Column('B', format='X', array=[[0], [1]])
        c3 = Column('C', format='4I', dim='(2, 2)',
                    array=[[0, 1, 2, 3], [4, 5, 6, 7]])
        c4 = Column('D', format='J', bscale=2.0, array=[0, 1])
        c5 = Column('E', format='A3', array=['abc', 'def'])
        c6 = Column('F', format='E', unit='m', array=[0.0, 1.0])
        c7 = Column('G', format='D', bzero=-0.1, array=[0.0, 1.0])
        c8 = Column('H', format='C', array=[0.0+1.0j, 2.0+3.0j])
        c9 = Column('I', format='M', array=[4.0+5.0j, 6.0+7.0j])
        c10 = Column('J', format='PI(2)', array=[[0, 1], [2, 3]])

        columns = [c1, c2, c3, c4, c5, c6, c7, c8, c9, c10]

        ta = BinTableHDU.from_columns(columns)
        tb = BinTableHDU.from_columns([c.copy() for c in columns])

        diff = TableDataDiff(ta.data, tb.data)
        assert diff.identical
        assert len(diff.common_columns) == 10
        assert diff.common_column_names == set('abcdefghij')
        assert diff.diff_ratio == 0
        assert diff.diff_total == 0
示例#11
0
    def test_diff_empty_tables(self):
        """
        Regression test for https://aeon.stsci.edu/ssb/trac/pyfits/ticket/178

        Ensure that diffing tables containing empty data doesn't crash.
        """

        c1 = Column('D', format='J')
        c2 = Column('E', format='J')
        thdu = BinTableHDU.from_columns([c1, c2], nrows=0)

        hdula = fits.HDUList([thdu])
        hdulb = fits.HDUList([thdu])

        diff = FITSDiff(hdula, hdulb)
        assert diff.identical
示例#12
0
    def test_diff_empty_tables(self):
        """
        Regression test for https://aeon.stsci.edu/ssb/trac/pyfits/ticket/178

        Ensure that diffing tables containing empty data doesn't crash.
        """

        c1 = Column('D', format='J')
        c2 = Column('E', format='J')
        thdu = BinTableHDU.from_columns([c1, c2], nrows=0)

        hdula = fits.HDUList([thdu])
        hdulb = fits.HDUList([thdu])

        diff = FITSDiff(hdula, hdulb)
        assert diff.identical
示例#13
0
    def test_different_table_data(self):
        """
        Test diffing table data on columns of several different data formats
        and dimensions.
        """

        ca1 = Column('A', format='L', array=[True, False])
        ca2 = Column('B', format='X', array=[[0], [1]])
        ca3 = Column('C',
                     format='4I',
                     dim='(2, 2)',
                     array=[[0, 1, 2, 3], [4, 5, 6, 7]])
        ca4 = Column('D', format='J', bscale=2.0, array=[0.0, 2.0])
        ca5 = Column('E', format='A3', array=['abc', 'def'])
        ca6 = Column('F', format='E', unit='m', array=[0.0, 1.0])
        ca7 = Column('G', format='D', bzero=-0.1, array=[0.0, 1.0])
        ca8 = Column('H', format='C', array=[0.0 + 1.0j, 2.0 + 3.0j])
        ca9 = Column('I', format='M', array=[4.0 + 5.0j, 6.0 + 7.0j])
        ca10 = Column('J', format='PI(2)', array=[[0, 1], [2, 3]])

        cb1 = Column('A', format='L', array=[False, False])
        cb2 = Column('B', format='X', array=[[0], [0]])
        cb3 = Column('C',
                     format='4I',
                     dim='(2, 2)',
                     array=[[0, 1, 2, 3], [5, 6, 7, 8]])
        cb4 = Column('D', format='J', bscale=2.0, array=[2.0, 2.0])
        cb5 = Column('E', format='A3', array=['abc', 'ghi'])
        cb6 = Column('F', format='E', unit='m', array=[1.0, 2.0])
        cb7 = Column('G', format='D', bzero=-0.1, array=[2.0, 3.0])
        cb8 = Column('H', format='C', array=[1.0 + 1.0j, 2.0 + 3.0j])
        cb9 = Column('I', format='M', array=[5.0 + 5.0j, 6.0 + 7.0j])
        cb10 = Column('J', format='PI(2)', array=[[1, 2], [3, 4]])

        ta = BinTableHDU.from_columns(
            [ca1, ca2, ca3, ca4, ca5, ca6, ca7, ca8, ca9, ca10])
        tb = BinTableHDU.from_columns(
            [cb1, cb2, cb3, cb4, cb5, cb6, cb7, cb8, cb9, cb10])

        diff = TableDataDiff(ta.data, tb.data, numdiffs=20)
        assert not diff.identical
        # The column definitions are the same, but not the column values
        assert diff.diff_columns == ()
        assert diff.diff_values[0] == (('A', 0), (True, False))
        assert diff.diff_values[1] == (('B', 1), ([1], [0]))
        assert diff.diff_values[2][0] == ('C', 1)
        assert (diff.diff_values[2][1][0] == [[4, 5], [6, 7]]).all()
        assert (diff.diff_values[2][1][1] == [[5, 6], [7, 8]]).all()
        assert diff.diff_values[3] == (('D', 0), (0, 2.0))
        assert diff.diff_values[4] == (('E', 1), ('def', 'ghi'))
        assert diff.diff_values[5] == (('F', 0), (0.0, 1.0))
        assert diff.diff_values[6] == (('F', 1), (1.0, 2.0))
        assert diff.diff_values[7] == (('G', 0), (0.0, 2.0))
        assert diff.diff_values[8] == (('G', 1), (1.0, 3.0))
        assert diff.diff_values[9] == (('H', 0), (0.0 + 1.0j, 1.0 + 1.0j))
        assert diff.diff_values[10] == (('I', 0), (4.0 + 5.0j, 5.0 + 5.0j))
        assert diff.diff_values[11][0] == ('J', 0)
        assert (diff.diff_values[11][1][0] == [0, 1]).all()
        assert (diff.diff_values[11][1][1] == [1, 2]).all()
        assert diff.diff_values[12][0] == ('J', 1)
        assert (diff.diff_values[12][1][0] == [2, 3]).all()
        assert (diff.diff_values[12][1][1] == [3, 4]).all()

        assert diff.diff_total == 13
        assert diff.diff_ratio == 0.65

        report = diff.report()
        assert ('Column A data differs in row 0:\n'
                '    a> True\n'
                '    b> False') in report
        assert ('...and at 1 more indices.\n'
                ' Column D data differs in row 0:') in report
        assert ('13 different table data element(s) found (65.00% different)'
                in report)
        assert report.count('more indices') == 1
示例#14
0
    def test_different_table_data(self):
        """
        Test diffing table data on columns of several different data formats
        and dimensions.
        """

        ca1 = Column('A', format='L', array=[True, False])
        ca2 = Column('B', format='X', array=[[0], [1]])
        ca3 = Column('C', format='4I', dim='(2, 2)',
                     array=[[0, 1, 2, 3], [4, 5, 6, 7]])
        ca4 = Column('D', format='J', bscale=2.0, array=[0.0, 2.0])
        ca5 = Column('E', format='A3', array=['abc', 'def'])
        ca6 = Column('F', format='E', unit='m', array=[0.0, 1.0])
        ca7 = Column('G', format='D', bzero=-0.1, array=[0.0, 1.0])
        ca8 = Column('H', format='C', array=[0.0+1.0j, 2.0+3.0j])
        ca9 = Column('I', format='M', array=[4.0+5.0j, 6.0+7.0j])
        ca10 = Column('J', format='PI(2)', array=[[0, 1], [2, 3]])

        cb1 = Column('A', format='L', array=[False, False])
        cb2 = Column('B', format='X', array=[[0], [0]])
        cb3 = Column('C', format='4I', dim='(2, 2)',
                     array=[[0, 1, 2, 3], [5, 6, 7, 8]])
        cb4 = Column('D', format='J', bscale=2.0, array=[2.0, 2.0])
        cb5 = Column('E', format='A3', array=['abc', 'ghi'])
        cb6 = Column('F', format='E', unit='m', array=[1.0, 2.0])
        cb7 = Column('G', format='D', bzero=-0.1, array=[2.0, 3.0])
        cb8 = Column('H', format='C', array=[1.0+1.0j, 2.0+3.0j])
        cb9 = Column('I', format='M', array=[5.0+5.0j, 6.0+7.0j])
        cb10 = Column('J', format='PI(2)', array=[[1, 2], [3, 4]])

        ta = BinTableHDU.from_columns([ca1, ca2, ca3, ca4, ca5, ca6, ca7,
                                       ca8, ca9, ca10])
        tb = BinTableHDU.from_columns([cb1, cb2, cb3, cb4, cb5, cb6, cb7,
                                       cb8, cb9, cb10])

        diff = TableDataDiff(ta.data, tb.data, numdiffs=20)
        assert not diff.identical
        # The column definitions are the same, but not the column values
        assert diff.diff_columns == ()
        assert diff.diff_values[0] == (('A', 0), (True, False))
        assert diff.diff_values[1] == (('B', 1), ([1], [0]))
        assert diff.diff_values[2][0] == ('C', 1)
        assert (diff.diff_values[2][1][0] == [[4, 5], [6, 7]]).all()
        assert (diff.diff_values[2][1][1] == [[5, 6], [7, 8]]).all()
        assert diff.diff_values[3] == (('D', 0), (0, 2.0))
        assert diff.diff_values[4] == (('E', 1), ('def', 'ghi'))
        assert diff.diff_values[5] == (('F', 0), (0.0, 1.0))
        assert diff.diff_values[6] == (('F', 1), (1.0, 2.0))
        assert diff.diff_values[7] == (('G', 0), (0.0, 2.0))
        assert diff.diff_values[8] == (('G', 1), (1.0, 3.0))
        assert diff.diff_values[9] == (('H', 0), (0.0+1.0j, 1.0+1.0j))
        assert diff.diff_values[10] == (('I', 0), (4.0+5.0j, 5.0+5.0j))
        assert diff.diff_values[11][0] == ('J', 0)
        assert (diff.diff_values[11][1][0] == [0, 1]).all()
        assert (diff.diff_values[11][1][1] == [1, 2]).all()
        assert diff.diff_values[12][0] == ('J', 1)
        assert (diff.diff_values[12][1][0] == [2, 3]).all()
        assert (diff.diff_values[12][1][1] == [3, 4]).all()

        assert diff.diff_total == 13
        assert diff.diff_ratio == 0.65

        report = diff.report()
        assert ('Column A data differs in row 0:\n'
                '    a> True\n'
                '    b> False') in report
        assert ('...and at 1 more indices.\n'
                ' Column D data differs in row 0:') in report
        assert ('13 different table data element(s) found (65.00% different)'
                in report)
        assert report.count('more indices') == 1