def test_difference():
"Test table.difference"
ds = datasets.get_uv()
print table.difference("fltvar", "A", "a1", "a2", "rm", ds=ds)
print table.difference("fltvar", "A", "a1", "a2", "rm", by="B", ds=ds)
print table.difference("fltvar", "A%B", ("a1", "b1"), ("a2", "b2"), "rm", ds=ds)
# create bigger dataset
ds["C", :] = "c1"
ds2 = datasets.get_uv()
ds2["C", :] = "c2"
ds = combine((ds, ds2))
print table.difference("fltvar", "A", "a1", "a2", "rm", "B%C", ds=ds)
print table.difference("fltvar", "A%B", ("a1", "b1"), ("a2", "b2"), "rm", "C", ds=ds)
def test_dataset_combining():
"Test Dataset combination methods"
ds = datasets.get_uv()
del ds['fltvar'], ds['intvar'], ds['A']
ds2 = datasets.get_uv()
del ds2['fltvar'], ds2['intvar']
ds.update(ds2)
assert_array_equal(ds['A'], ds2['A'])
ds2 = datasets.get_uv()
del ds2['fltvar'], ds2['intvar']
ds2['B'][5] = 'something_else'
del ds['A']
assert_raises(ValueError, ds.update, ds2)
def test_dataset_combining():
"Test Dataset combination methods"
ds = datasets.get_uv()
del ds['fltvar'], ds['intvar'], ds['A']
ds2 = datasets.get_uv()
del ds2['fltvar'], ds2['intvar']
ds.update(ds2)
assert_array_equal(ds['A'], ds2['A'])
ds2 = datasets.get_uv()
del ds2['fltvar'], ds2['intvar']
ds2['B'][5] = 'something_else'
del ds['A']
assert_raises(ValueError, ds.update, ds2)
def test_barplot():
"Test plot.Barplot"
ds = datasets.get_uv()
# average
plot.Barplot('fltvar', ds=ds)
plot.Barplot('fltvar', match='rm', ds=ds)
# one category
p = plot.Barplot('fltvar + 2', 'A', ds=ds)
assert p.get_ylim()[0] == 0
p = plot.Barplot('fltvar - 2', 'A', 'rm', ds=ds)
assert p.get_ylim()[1] == 0
# multiple categories
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds)
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, pool_error=False)
# test against 0
plot.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds)
# cells
plot.Barplot('fltvar', 'A%B', cells=(('a2', 'b2'), ('a1', 'b1')), ds=ds)
plot.Barplot('fltvar',
'A%B',
match='rm',
cells=(('a2', 'b2'), ('a1', 'b1')),
ds=ds)
# Fixed top
p = plot.Barplot('fltvar', 'A%B', ds=ds, top=2, test_markers=False)
ax = p._axes[0]
assert ax.get_ylim()[1] == 2
def test_difference():
"Test table.difference"
ds = datasets.get_uv()
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', ds=ds)
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', by='B', ds=ds)
print table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm',
ds=ds)
# create bigger dataset
ds['C', :] = 'c1'
ds2 = datasets.get_uv()
ds2['C', :] = 'c2'
ds = combine((ds, ds2))
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', 'B%C', ds=ds)
print table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm',
'C', ds=ds)
def test_barplot():
"Test plot.Barplot"
ds = datasets.get_uv()
# one category
plot.Barplot('fltvar', ds=ds, test=False, show=False)
plot.Barplot('fltvar', ds=ds, show=False)
plot.Barplot('fltvar', match='rm', ds=ds, show=False)
# multiple categories
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, show=False)
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, pool_error=False,
show=False)
plot.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds, show=False)
# cells
plot.Barplot('fltvar', 'A%B', cells=(('a2', 'b2'), ('a1', 'b1')), ds=ds, show=False)
plot.Barplot('fltvar', 'A%B', match='rm', cells=(('a2', 'b2'), ('a1', 'b1')),
ds=ds, show=False)
# Fixed top
p = plot.Barplot('fltvar', 'A%B', ds=ds, top=2, test_markers=False,
show=False)
ax = p._axes[0]
eq_(ax.get_ylim()[1], 2)
def test_regression():
"Test plot.Regression"
ds = datasets.get_uv()
plot.Regression('fltvar', 'fltvar2', ds=ds)
plot.Regression('fltvar', 'fltvar2', 'A', ds=ds)
plot.Regression('fltvar', 'fltvar2', 'A%B', ds=ds)
def test_barplot():
"Test plot.uv.barplot"
plot.configure_backend(False, False)
ds = datasets.get_uv()
plot.uv.Barplot('fltvar', 'A%B', match='rm', ds=ds)
plot.uv.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds)
plt.close('all')
def test_generate_colors():
"Test functions for assigning colors to cells"
ds = datasets.get_uv()
cells_1 = ('A', 'B')
cells_2 = ('a', 'b', 'c')
colors = plot.colors_for_oneway(cells_1)
assert len(colors) == len(cells_1)
colors = plot.colors_for_twoway(cells_1, cells_2)
assert len(colors) == len(cells_1) * len(cells_2)
colors = plot.colors_for_twoway(cells_2, cells_1)
assert len(colors) == len(cells_1) * len(cells_2)
# colors_for_categorial()
f = ds['A']
colors = plot.colors_for_categorial(f)
for cell in f.cells:
assert cell in colors
i = ds.eval("A%B")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert cell in colors
i = ds.eval("A%B%rm")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert cell in colors
with pytest.raises(TypeError):
plot.colors_for_categorial("A%B")
def test_cast_to_ndvar():
"Test table.cast_to_ndvar()"
long_ds = datasets.get_uv()
long_ds['scalar'] = long_ds['A'] == 'a2'
long_ds['time'] = long_ds.eval('A%B').as_var({
('a1', 'b1'): 0.,
('a1', 'b2'): 0.1,
('a2', 'b1'): 0.2,
('a2', 'b2'): 0.3,
})
# categorial
ds = table.cast_to_ndvar('fltvar', 'A', 'B%rm', ds=long_ds, name='new')
assert ds.n_cases == long_ds.n_cases / 2
assert ds['new'].A == Categorial('A', ('a1', 'a2'))
# scalar
ds2 = table.cast_to_ndvar('fltvar', 'scalar', 'B%rm', ds=long_ds, dim='newdim', name='new')
assert ds2.n_cases == long_ds.n_cases / 2
assert ds2['new'].newdim == Scalar('newdim', [False, True])
assert_array_equal(ds['new'].x, ds2['new'].x)
# time
ds = table.cast_to_ndvar('fltvar', 'time', 'rm', ds=long_ds, dim='uts', name='y')
assert ds.n_cases == long_ds.n_cases / 4
assert ds['y'].time == UTS(0, 0.1, 4)
def test_report():
"Test fmtxt.Report class"
tempdir = tempfile.mkdtemp()
report = fmtxt.Report("Test Report")
section = report.add_section('unicode: \xe2 abc')
ds = datasets.get_uv()
p = plot.Barplot('fltvar', 'A', sub="B=='b1'", ds=ds, show=False)
image = p.image()
section.add_figure("test", image)
report.sign()
# report output
print(report)
dst = os.path.join(tempdir, 'report.html')
report.save_html(dst)
# test meta attribute reading
eq_(read_meta(dst), {})
report.save_html(dst, meta={'samples': 100, 'text': 'blah'})
eq_(read_meta(dst), {'samples': '100', 'text': 'blah'})
# clean up
shutil.rmtree(tempdir)
def test_align():
"Testing align() and align1() functions"
ds = datasets.get_uv()
ds.index()
idx4 = np.arange(0, ds.n_cases, 4)
idx4i = idx4[::-1]
ds2 = ds.sub(np.arange(0, ds.n_cases, 2))
# align1: align Dataset to index
dsa = align1(ds2, idx4)
assert_array_equal(dsa['index'], idx4, "align1() failure")
dsa = align1(ds2, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
# d_idx as Var
dsa = align1(ds2[::2], idx4, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
assert_raises(ValueError, align1, ds2, idx4, idx4i)
# Factor index
assert_raises(ValueError, align1, ds, ds['rm', ::-1], 'rm')
fds = ds[:20]
dsa = align1(fds, fds['rm', ::-1], 'rm')
assert_array_equal(dsa['index'], np.arange(19, -1, -1), "align1 Factor")
# align two datasets
dsa1, dsa2 = align(ds, ds2)
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
dsa1, dsa2 = align(ds, ds2[::-1])
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
def test_interaction():
"Test Interaction"
ds = datasets.get_uv()
A = ds['A']
B = ds['B']
i = A % B
# eq for sequence
assert_array_equal(i == A % B, True)
assert_array_equal(i == B % A, False)
assert_array_equal(i == A, False)
assert_array_equal(i == ds['fltvar'], False)
assert_array_equal(ds.eval("A%B") == Factor(ds['A']) % B, True)
# eq for element
for a, b in product(A.cells, B.cells):
assert_array_equal(i == (a, b), np.logical_and(A == a, B == b))
# Interaction.as_factor()
a = Factor('aabb')
i = a % Factor('cdcd')
assert_dataobj_equal(i.as_factor(), Factor(['a c', 'a d', 'b c', 'b d']))
i = a % Factor(['c', '', 'c', ''])
assert_dataobj_equal(i.as_factor(), Factor(['a c', 'a', 'b c', 'b']))
# pickling
ip = pickle.loads(pickle.dumps(i))
assert_dataobj_equal(ip, i)
def test_barplot():
"Test plot.Barplot"
ds = datasets.get_uv()
# one category
plot.Barplot('fltvar', ds=ds, test=False, show=False)
plot.Barplot('fltvar', ds=ds, show=False)
plot.Barplot('fltvar', match='rm', ds=ds, show=False)
# multiple categories
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, show=False)
plot.Barplot('fltvar',
'A%B',
match='rm',
ds=ds,
pool_error=False,
show=False)
plot.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds, show=False)
# Fixed top
p = plot.Barplot('fltvar',
'A%B',
ds=ds,
top=2,
test_markers=False,
show=False)
ax = p._axes[0]
eq_(ax.get_ylim()[1], 2)
def test_tsv_io():
"""Test tsv I/O"""
names = ['A', 'B', 'rm', 'intvar', 'fltvar', 'fltvar2', 'index']
# get Dataset
ds = datasets.get_uv()
ds['fltvar'][5:10] = np.nan
ds[:4, 'rm'] = ''
# save and load
tempdir = tempfile.mkdtemp()
try:
dst = os.path.join(tempdir, 'ds.txt')
ds.save_txt(dst)
ds1 = load.tsv(dst)
ds2 = load.tsv(dst, skiprows=1, names=names)
assert_dataset_equal(ds1, ds, "TSV write/read test failed", 10)
assert_dataset_equal(ds2, ds, "TSV write/read test failed", 10)
# guess data types with missing
intvar2 = ds['intvar'].as_factor()
intvar2[10:] = ''
ds_intvar = Dataset((intvar2, ))
ds_intvar.save_txt(dst)
ds_intvar1 = load.tsv(dst, empty='nan')
assert_dataobj_equal(ds_intvar1['intvar', :10], ds['intvar', :10])
assert_array_equal(ds_intvar1['intvar', 10:], np.nan)
finally:
shutil.rmtree(tempdir)
def test_boxplot():
"Test plot.Boxplot"
ds = datasets.get_uv()
plot.Boxplot('fltvar', 'A%B', match='rm', ds=ds)
# one category
plot.Boxplot('fltvar', ds=ds, test=False)
plot.Boxplot('fltvar', ds=ds)
plot.Boxplot('fltvar', match='rm', ds=ds)
# cells
plot.Boxplot('fltvar', 'A%B', cells=(('a2', 'b2'), ('a1', 'b1')), ds=ds)
plot.Boxplot('fltvar', 'A%B', match='rm', cells=(('a2', 'b2'), ('a1', 'b1')), ds=ds)
# many pairwise significances
ds['fltvar'][ds.eval("A%B==('a1','b1')")] += 1
ds['fltvar'][ds.eval("A%B==('a2','b2')")] -= 1
ds['C'] = Factor('qw', repeat=10, tile=4)
plot.Boxplot('fltvar', 'A%B%C', ds=ds)
# long labels
ds['A'].update_labels({'a1': 'a very long label', 'a2': 'another very long label'})
p = plot.Barplot('fltvar', 'A%B', ds=ds)
labels = p._ax.get_xticklabels()
bbs = [l.get_window_extent() for l in labels]
for i in range(len(bbs) - 1):
assert bbs[i].x1 < bbs[i + 1].x0
def test_align():
"Testing align() and align1() functions"
ds = datasets.get_uv()
ds.index()
idx4 = np.arange(0, ds.n_cases, 4)
idx4i = idx4[::-1]
ds2 = ds.sub(np.arange(0, ds.n_cases, 2))
# align1: align Dataset to index
dsa = align1(ds2, idx4)
assert_array_equal(dsa['index'], idx4, "align1() failure")
dsa = align1(ds2, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
# d_idx as Var
dsa = align1(ds2[::2], idx4, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
assert_raises(ValueError, align1, ds2, idx4, idx4i)
# Factor index
assert_raises(ValueError, align1, ds, ds['rm', ::-1], 'rm')
fds = ds[:20]
dsa = align1(fds, fds['rm', ::-1], 'rm')
assert_array_equal(dsa['index'], np.arange(19, -1, -1), "align1 Factor")
# align two datasets
dsa1, dsa2 = align(ds, ds2)
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
dsa1, dsa2 = align(ds, ds2[::-1])
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
def test_correlations():
"Test test.correlations()"
ds = datasets.get_uv()
res = test.correlations('fltvar', 'fltvar2', ds=ds)
print(res)
eq_(str(res[2][0]).strip(), '.398')
res = test.correlations('fltvar', 'fltvar2', ds=ds, asds=True)
assert_almost_equal(res[0, 'r'], .398, 3)
res = test.correlations('fltvar', 'fltvar2', 'A', ds=ds)
print(res)
eq_(str(res[2][0]).strip(), 'a1')
eq_(str(res[2][1]).strip(), '-0.149')
eq_(str(res[3][1]).strip(), '.740')
res = test.correlations('fltvar', 'fltvar2', 'A', ds=ds, asds=True)
eq_(res[0, 'A'], 'a1')
assert_almost_equal(res[0, 'r'], -0.149, 3)
assert_almost_equal(res[1, 'r'], .740, 3)
res = test.correlations('fltvar', 'fltvar2', 'A%B', ds=ds)
print(res)
eq_(str(res[2][2]).strip(), '-0.276')
res = test.correlations('fltvar', 'fltvar2', 'A%B', ds=ds, asds=True)
assert_almost_equal(res[0, 'r'], -0.276, 3)
res = test.correlations('fltvar', ('fltvar2', 'intvar'), 'A%B', ds=ds)
print(res)
eq_(str(res[2][1]).strip(), 'a1')
eq_(str(res[2][2]).strip(), 'b1')
eq_(str(res[2][3]).strip(), '-0.276')
res = test.correlations('fltvar', ('fltvar2', 'intvar'), 'A%B', ds=ds, asds=True)
assert_almost_equal(res[0, 'r'], -0.276, 3)
def test_correlation():
"Test plot.Correlation()"
ds = datasets.get_uv()
plot.Correlation('fltvar', 'fltvar2', ds=ds)
plot.Correlation('fltvar', 'fltvar2', 'A', ds=ds)
plot.Correlation('fltvar', 'fltvar2', 'A%B', ds=ds)
def test_boxplot():
"Test plot.Boxplot"
ds = datasets.get_uv()
plot.Boxplot('fltvar', 'A%B', match='rm', ds=ds, show=False)
# one category
plot.Boxplot('fltvar', ds=ds, test=False, show=False)
plot.Boxplot('fltvar', ds=ds, show=False)
plot.Boxplot('fltvar', match='rm', ds=ds, show=False)
# many pairwise significances
ds['fltvar'][ds.eval("A%B==('a1','b1')")] += 1
ds['fltvar'][ds.eval("A%B==('a2','b2')")] -= 1
ds['C'] = Factor('qw', repeat=10, tile=4)
plot.Boxplot('fltvar', 'A%B%C', ds=ds, show=False)
# long labels
ds['A'].relabel({
'a1': 'a very long label',
'a2': 'another very long label'
})
p = plot.Barplot('fltvar', 'A%B', ds=ds, show=False)
labels = p._ax.get_xticklabels()
bbs = [l.get_window_extent() for l in labels]
for i in xrange(len(bbs) - 1):
ok_(bbs[i].x1 < bbs[i + 1].x0)
def test_difference():
"Test table.difference"
ds = datasets.get_uv()
# add a variables that should stay in the dataset
labels = {c: c[-1] for c in ds['rm'].cells}
ds['rmf'] = Factor(ds['rm'], labels=labels)
dds = table.difference('fltvar', 'A', 'a1', 'a2', 'rm', ds=ds)
assert repr(dds) == "<Dataset n_cases=20 {'rm':F, 'fltvar':V, 'rmf':F}>"
assert_array_equal(dds['rmf'], Factor(dds['rm'], labels=labels))
dds = table.difference('fltvar', 'A', 'a1', 'a2', 'rm', by='B', ds=ds)
assert repr(dds) == "<Dataset n_cases=40 {'rm':F, 'fltvar':V, 'B':F, 'rmf':F}>"
# difference of the difference
ddds = table.difference('fltvar', 'B', 'b1', 'b2', 'rm', ds=dds)
assert repr(ddds) == "<Dataset n_cases=20 {'rm':F, 'fltvar':V, 'rmf':F}>"
dds = table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm', ds=ds)
assert repr(dds) == "<Dataset n_cases=20 {'rm':F, 'fltvar':V, 'rmf':F}>"
# create bigger dataset
ds2 = ds.copy()
ds['C', :] = 'c1'
ds2['C', :] = 'c2'
ds = combine((ds, ds2))
dds = table.difference('fltvar', 'A', 'a1', 'a2', 'rm', by='B%C', ds=ds)
assert repr(dds) == "<Dataset n_cases=80 {'rm':F, 'fltvar':V, 'B':F, 'C':F, 'rmf':F}>"
dds = table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm', by='C', ds=ds)
assert repr(dds) == "<Dataset n_cases=40 {'rm':F, 'fltvar':V, 'C':F, 'rmf':F}>"
def test_cast_to_ndvar():
"Test table.cast_to_ndvar()"
long_ds = datasets.get_uv()
long_ds['scalar'] = long_ds['A'] == 'a2'
long_ds['time'] = long_ds.eval('A%B').as_var({
('a1', 'b1'): 0.,
('a1', 'b2'): 0.1,
('a2', 'b1'): 0.2,
('a2', 'b2'): 0.3,
})
# categorial
ds = table.cast_to_ndvar('fltvar', 'A', 'B%rm', ds=long_ds, name='new')
assert ds.n_cases == long_ds.n_cases / 2
assert ds['new'].A == Categorial('A', ('a1', 'a2'))
# scalar
ds2 = table.cast_to_ndvar('fltvar', 'scalar', 'B%rm', ds=long_ds, dim='newdim', name='new')
assert ds2.n_cases == long_ds.n_cases / 2
assert ds2['new'].newdim == Scalar('newdim', [False, True])
assert_array_equal(ds['new'].x, ds2['new'].x)
# time
ds = table.cast_to_ndvar('fltvar', 'time', 'rm', ds=long_ds, dim='uts', name='y')
assert ds.n_cases == long_ds.n_cases / 4
assert ds['y'].time == UTS(0, 0.1, 4)
def test_correlations():
"Test test.correlations()"
ds = datasets.get_uv()
res = test.correlations("fltvar", "fltvar2", ds=ds)
print(res)
eq_(str(res[2][0]).strip(), ".398")
res = test.correlations("fltvar", "fltvar2", ds=ds, asds=True)
assert_almost_equal(res[0, "r"], 0.398, 3)
res = test.correlations("fltvar", "fltvar2", "A", ds=ds)
print(res)
eq_(str(res[2][0]).strip(), "a1")
eq_(str(res[2][1]).strip(), "-0.149")
eq_(str(res[3][1]).strip(), ".740")
res = test.correlations("fltvar", "fltvar2", "A", ds=ds, asds=True)
eq_(res[0, "A"], "a1")
assert_almost_equal(res[0, "r"], -0.149, 3)
assert_almost_equal(res[1, "r"], 0.740, 3)
res = test.correlations("fltvar", "fltvar2", "A%B", ds=ds)
print(res)
eq_(str(res[2][2]).strip(), "-0.276")
res = test.correlations("fltvar", "fltvar2", "A%B", ds=ds, asds=True)
assert_almost_equal(res[0, "r"], -0.276, 3)
res = test.correlations("fltvar", ("fltvar2", "intvar"), "A%B", ds=ds)
print(res)
eq_(str(res[2][1]).strip(), "a1")
eq_(str(res[2][2]).strip(), "b1")
eq_(str(res[2][3]).strip(), "-0.276")
res = test.correlations("fltvar", ("fltvar2", "intvar"), "A%B", ds=ds, asds=True)
assert_almost_equal(res[0, "r"], -0.276, 3)
def test_correlation():
"Test plot.Correlation()"
ds = datasets.get_uv()
plot.Correlation('fltvar', 'fltvar2', ds=ds, show=False)
plot.Correlation('fltvar', 'fltvar2', 'A', ds=ds, show=False)
plot.Correlation('fltvar', 'fltvar2', 'A%B', ds=ds, show=False)
def test_barplot():
"Test plot.uv.barplot"
plot.configure_backend(False, False)
ds = datasets.get_uv()
plot.uv.Barplot('fltvar', 'A%B', match='rm', ds=ds)
plot.uv.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds)
plt.close('all')
def test_report():
"Test fmtxt.Report class"
tempdir = tempfile.mkdtemp()
report = fmtxt.Report("Test Report")
section = report.add_section(u'unicode: \xe2 abc')
ds = datasets.get_uv()
p = plot.Barplot('fltvar', 'A', sub="B=='b1'", ds=ds, show=False)
image = p.image()
section.add_figure("test", image)
report.sign()
# report output
print(report)
dst = os.path.join(tempdir, 'report.html')
report.save_html(dst)
# test meta attribute reading
eq_(read_meta(dst), {})
report.save_html(dst, meta={'samples': 100, 'text': 'blah'})
eq_(read_meta(dst), {'samples': '100', 'text': 'blah'})
# clean up
shutil.rmtree(tempdir)
def test_regression():
"Test plot.Regression"
ds = datasets.get_uv()
plot.Regression('fltvar', 'fltvar2', ds=ds, show=False)
plot.Regression('fltvar', 'fltvar2', 'A', ds=ds, show=False)
plot.Regression('fltvar', 'fltvar2', 'A%B', ds=ds, show=False)
def test_anova():
"Test ANOVA"
from rpy2.robjects import r
r_require('car')
ds = datasets.get_uv()
ds.to_r('ds')
# fixed effects
aov = test.anova('fltvar', 'A*B', ds=ds)
print aov
fs = run_on_lm_fitter('fltvar', 'A*B', ds)
r_res = r("Anova(lm(fltvar ~ A * B, ds, type=2))")
assert_f_tests_equal(aov.f_tests, r_res, fs, 'Anova')
# random effects
aov = test.anova('fltvar', 'A*B*rm', ds=ds)
print aov
fs = run_on_lm_fitter('fltvar', 'A*B*rm', ds)
r('test.aov <- aov(fltvar ~ A * B + Error(rm / (A * B)), ds)')
print r('test.summary <- summary(test.aov)')
r_res = r['test.summary'][1:]
assert_f_tests_equal(aov.f_tests, r_res, fs, 'rmaov')
# not fully specified model with random effects
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*rm', ds=ds)
# empty cells
dss = ds.sub("A%B != ('a2', 'b2')")
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*B', ds=dss)
assert_raises(NotImplementedError, run_on_lm_fitter, 'fltvar', 'A*B', ds=dss)
dss = ds.sub("A%B != ('a1', 'b1')")
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*B', ds=dss)
assert_raises(NotImplementedError, run_on_lm_fitter, 'fltvar', 'A*B', ds=dss)
def test_generate_colors():
"Test functions for assigning colors to cells"
ds = datasets.get_uv()
cells_1 = ('A', 'B')
cells_2 = ('a', 'b', 'c')
colors = plot.colors_for_oneway(cells_1)
eq_(len(colors), len(cells_1))
colors = plot.colors_for_twoway(cells_1, cells_2)
eq_(len(colors), len(cells_1) * len(cells_2))
colors = plot.colors_for_twoway(cells_2, cells_1)
eq_(len(colors), len(cells_1) * len(cells_2))
# colors_for_categorial()
f = ds['A']
colors = plot.colors_for_categorial(f)
for cell in f.cells:
assert_in(cell, colors)
i = ds.eval("A%B")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert_in(cell, colors)
i = ds.eval("A%B%rm")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert_in(cell, colors)
assert_raises(TypeError, plot.colors_for_categorial, "A%B")
def test_difference():
"Test tabe.difference"
ds = datasets.get_uv()
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', ds=ds)
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', by='B', ds=ds)
print table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm',
ds=ds)
# create bigger dataset
ds['C', :] = 'c1'
ds2 = datasets.get_uv()
ds2['C', :] = 'c2'
ds = combine((ds, ds2))
print table.difference('fltvar', 'A', 'a1', 'a2', 'rm', 'B%C', ds=ds)
print table.difference('fltvar', 'A%B', ('a1', 'b1'), ('a2', 'b2'), 'rm',
'C', ds=ds)
def test_interaction():
"Test Interaction"
ds = datasets.get_uv()
A = ds['A']
B = ds['B']
i = A % B
# eq for sequence
assert_array_equal(i == A % B, True)
assert_array_equal(i == B % A, False)
assert_array_equal(i == A, False)
assert_array_equal(i == ds['fltvar'], False)
assert_array_equal(ds.eval("A%B") == Factor(ds['A']) % B, True)
# eq for element
for a, b in product(A.cells, B.cells):
assert_array_equal(i == (a, b), np.logical_and(A == a, B == b))
# Interaction.as_factor()
a = Factor('aabb')
i = a % Factor('cdcd')
assert_dataobj_equal(i.as_factor(), Factor(['a c', 'a d', 'b c', 'b d']))
i = a % Factor(['c', '', 'c', ''])
assert_dataobj_equal(i.as_factor(), Factor(['a c', 'a', 'b c', 'b']))
# pickling
ip = pickle.loads(pickle.dumps(i))
assert_dataobj_equal(ip, i)
def test_tsv_io():
"""Test tsv I/O"""
names = ['A', 'B', 'rm', 'intvar', 'fltvar', 'fltvar2', 'index']
# get Dataset
ds = datasets.get_uv()
ds['fltvar'][5:10] = np.nan
ds[:4, 'rm'] = ''
# save and load
tempdir = tempfile.mkdtemp()
try:
dst = os.path.join(tempdir, 'ds.txt')
ds.save_txt(dst)
ds1 = load.tsv(dst)
ds2 = load.tsv(dst, skiprows=1, names=names)
assert_dataset_equal(ds1, ds, "TSV write/read test failed", 10)
assert_dataset_equal(ds2, ds, "TSV write/read test failed", 10)
# guess data types with missing
intvar2 = ds['intvar'].as_factor()
intvar2[10:] = ''
ds_intvar = Dataset((intvar2,))
ds_intvar.save_txt(dst)
ds_intvar1 = load.tsv(dst, empty='nan')
assert_dataobj_equal(ds_intvar1['intvar', :10], ds['intvar', :10])
assert_array_equal(ds_intvar1['intvar', 10:], np.nan)
finally:
shutil.rmtree(tempdir)
def test_generate_colors():
"Test functions for assigning colors to cells"
ds = datasets.get_uv()
cells_1 = ('A', 'B')
cells_2 = ('a', 'b', 'c')
colors = plot.colors_for_oneway(cells_1)
assert_equal(len(colors), len(cells_1))
colors = plot.colors_for_twoway(cells_1, cells_2)
assert_equal(len(colors), len(cells_1) * len(cells_2))
colors = plot.colors_for_twoway(cells_2, cells_1)
assert_equal(len(colors), len(cells_1) * len(cells_2))
# colors_for_categorial()
f = ds['A']
colors = plot.colors_for_categorial(f)
for cell in f.cells:
assert_in(cell, colors)
i = ds.eval("A%B")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert_in(cell, colors)
i = ds.eval("A%B%rm")
colors = plot.colors_for_categorial(i)
for cell in i.cells:
assert_in(cell, colors)
assert_raises(TypeError, plot.colors_for_categorial, "A%B")
def test_anova_eq():
"Test ANOVA against r-ez"
r_require('ez')
ds = datasets.get_uv(nrm=True)
ds.to_r('ds')
# nested random effect
aov_explicit = test.ANOVA('fltvar', 'A + B + A%B + nrm(B) + A%nrm(B)', ds=ds)
aov = test.ANOVA('fltvar', 'A * B * nrm(B)', ds=ds)
assert str(aov_explicit) == str(aov)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A * B * nrm(B)', ds)
fnds = run_as_ndanova('fltvar', 'A * B * nrm(B)', ds)
r_res = r('ezANOVA(ds, fltvar, nrm, A, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# sub parameter
r1 = test.ANOVA('fltvar', 'B * rm', ds=ds.sub('A == "a1"'))
r2 = test.ANOVA('fltvar', 'B * rm', sub='A == "a1"', ds=ds)
assert str(r2) == str(r1)
# not fully specified model with random effects
with pytest.raises(IncompleteModel):
test.anova('fltvar', 'A*rm', ds=ds)
# unequal group size, 1-way
sds = ds.sub("A == 'a1'").sub("nrm.isnotin(('s037', 's038', 's039'))")
sds.to_r('sds')
aov = test.ANOVA('fltvar', 'B * nrm(B)', ds=sds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'B * nrm(B)', sds)
fnds = run_as_ndanova('fltvar', 'B * nrm(B)', sds)
with r_warning_filter: # type argument to ezANOVA
r_res = r('ezANOVA(sds, fltvar, nrm, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# unequal group size, 2-way
sds = ds.sub("nrm.isnotin(('s037', 's038', 's039'))")
sds.to_r('sds')
aov = test.ANOVA('fltvar', 'A * B * nrm(B)', ds=sds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A * B * nrm(B)', sds)
fnds = run_as_ndanova('fltvar', 'A * B * nrm(B)', sds)
with r_warning_filter: # type argument to ezANOVA
r_res = r('ezANOVA(sds, fltvar, nrm, A, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# empty cells
dss = ds.sub("A%B != ('a2', 'b2')")
with pytest.raises(NotImplementedError):
test.anova('fltvar', 'A*B', ds=dss)
with pytest.raises(NotImplementedError):
run_on_lm_fitter('fltvar', 'A*B', ds=dss)
dss = ds.sub("A%B != ('a1', 'b1')")
with pytest.raises(NotImplementedError):
test.anova('fltvar', 'A*B', ds=dss)
with pytest.raises(NotImplementedError):
run_on_lm_fitter('fltvar', 'A*B', ds=dss)
def test_anova_eq():
"Test ANOVA against r-ez"
r_require('ez')
ds = datasets.get_uv(nrm=True)
ds.to_r('ds')
# nested random effect
aov_explicit = test.ANOVA('fltvar', 'A + B + A%B + nrm(B) + A%nrm(B)', ds=ds)
aov = test.ANOVA('fltvar', 'A * B * nrm(B)', ds=ds)
assert str(aov_explicit) == str(aov)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A * B * nrm(B)', ds)
fnds = run_as_ndanova('fltvar', 'A * B * nrm(B)', ds)
r_res = r('ezANOVA(ds, fltvar, nrm, A, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# sub parameter
r1 = test.ANOVA('fltvar', 'B * rm', ds=ds.sub('A == "a1"'))
r2 = test.ANOVA('fltvar', 'B * rm', sub='A == "a1"', ds=ds)
assert str(r2) == str(r1)
# not fully specified model with random effects
with pytest.raises(IncompleteModel):
test.anova('fltvar', 'A*rm', ds=ds)
# unequal group size, 1-way
sds = ds.sub("A == 'a1'").sub("nrm.isnotin(('s037', 's038', 's039'))")
sds.to_r('sds')
aov = test.ANOVA('fltvar', 'B * nrm(B)', ds=sds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'B * nrm(B)', sds)
fnds = run_as_ndanova('fltvar', 'B * nrm(B)', sds)
with r_warning_filter: # type argument to ezANOVA
r_res = r('ezANOVA(sds, fltvar, nrm, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# unequal group size, 2-way
sds = ds.sub("nrm.isnotin(('s037', 's038', 's039'))")
sds.to_r('sds')
aov = test.ANOVA('fltvar', 'A * B * nrm(B)', ds=sds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A * B * nrm(B)', sds)
fnds = run_as_ndanova('fltvar', 'A * B * nrm(B)', sds)
with r_warning_filter: # type argument to ezANOVA
r_res = r('ezANOVA(sds, fltvar, nrm, A, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# empty cells
dss = ds.sub("A%B != ('a2', 'b2')")
with pytest.raises(NotImplementedError):
test.anova('fltvar', 'A*B', ds=dss)
with pytest.raises(NotImplementedError):
run_on_lm_fitter('fltvar', 'A*B', ds=dss)
dss = ds.sub("A%B != ('a1', 'b1')")
with pytest.raises(NotImplementedError):
test.anova('fltvar', 'A*B', ds=dss)
with pytest.raises(NotImplementedError):
run_on_lm_fitter('fltvar', 'A*B', ds=dss)
def test_ttest():
"Test test.ttest"
ds = datasets.get_uv()
print test.ttest('fltvar', ds=ds)
print test.ttest('fltvar', 'A', ds=ds)
print test.ttest('fltvar', 'A%B', ds=ds)
print test.ttest('fltvar', 'A', match='rm', ds=ds)
def test_ttest():
"""Test test.ttest()"""
ds = datasets.get_uv()
print(test.ttest('fltvar', ds=ds))
print(test.ttest('fltvar', 'A', ds=ds))
print(test.ttest('fltvar', 'A%B', ds=ds))
print(test.ttest('fltvar', 'A', match='rm', ds=ds))
print(test.ttest('fltvar', 'A', 'a1', match='rm', ds=ds))
print(test.ttest('fltvar', 'A%B', ('a1', 'b1'), match='rm', ds=ds))
def test_repmeas():
"Test table.repmeas (repeated measures table)"
ds = datasets.get_uv()
print table.repmeas("fltvar", "A", "rm", ds=ds)
print table.repmeas("fltvar", "A%B", "rm", ds=ds)
print table.repmeas("fltvar", "A", "B%rm", ds=ds)
# test naturalization of cellnames
ds["ANum"] = Factor(ds["A"], labels={"a1": "1", "a2": "2"})
print table.repmeas("fltvar", "ANum", "rm", ds=ds)
def test_repmeas():
"Test table.repmeas (repeated measures table)"
ds = datasets.get_uv()
print table.repmeas('fltvar', 'A', 'rm', ds=ds)
print table.repmeas('fltvar', 'A%B', 'rm', ds=ds)
print table.repmeas('fltvar', 'A', 'B%rm', ds=ds)
# test naturalization of cellnames
ds['ANum'] = Factor(ds['A'], labels={'a1': '1', 'a2': '2'})
print table.repmeas('fltvar', 'ANum', 'rm', ds=ds)
def test_nested_effects():
"""Test nested effects"""
ds = datasets.get_uv(nrm=True)
nested = ds.eval("nrm(B)")
eq_(nested.cells, ds['nrm'].cells)
# interaction
i = ds.eval("A % nrm(B)")
expected_cells = tuple((case['A'], case['nrm']) for case in ds.itercases())
def test_ttest():
"""Test test.ttest()"""
ds = datasets.get_uv()
print(test.ttest("fltvar", ds=ds))
print(test.ttest("fltvar", "A", ds=ds))
print(test.ttest("fltvar", "A%B", ds=ds))
print(test.ttest("fltvar", "A", match="rm", ds=ds))
print(test.ttest("fltvar", "A", "a1", match="rm", ds=ds))
print(test.ttest("fltvar", "A%B", ("a1", "b1"), match="rm", ds=ds))
def test_dataset_indexing():
"""Test Dataset indexing"""
ds = datasets.get_uv()
# indexing values
eq_(ds['A', 1], ds['A'][1])
eq_(ds[1, 'A'], ds['A'][1])
# indexing variables
assert_dataobj_equal(ds[:, 'A'], ds['A'])
assert_dataobj_equal(ds['A', :], ds['A'])
assert_dataobj_equal(ds[:10, 'A'], ds['A'][:10])
assert_dataobj_equal(ds['A', :10], ds['A'][:10])
# new Dataset through indexing
ds2 = Dataset()
ds2['A'] = ds['A']
assert_dataset_equal(ds[('A',)], ds2)
ds2['B'] = ds['B']
assert_dataset_equal(ds['A', 'B'], ds2)
assert_dataset_equal(ds[('A', 'B'), :10], ds2[:10])
assert_dataset_equal(ds[:10, ('A', 'B')], ds2[:10])
# assigning value
ds[2, 'A'] = 'hello'
eq_(ds[2, 'A'], 'hello')
ds['A', 2] = 'not_hello'
eq_(ds[2, 'A'], 'not_hello')
# assigning new factor
ds['C', :] = 'c'
ok_(np.all(ds.eval("C == 'c'")))
# assigning new Var
ds['D1', :] = 5.
ds[:, 'D2'] = 5.
assert_array_equal(ds['D1'], 5)
assert_array_equal(ds['D2'], 5)
# test illegal names
f = Factor('aaabbb')
assert_raises(ValueError, ds.__setitem__, '%dsa', f)
assert_raises(ValueError, ds.__setitem__, '432', f)
assert_raises(ValueError, ds.__setitem__, ('%dsa', slice(None)), 'value')
assert_raises(ValueError, ds.__setitem__, (slice(None), '%dsa'), 'value')
assert_raises(ValueError, ds.__setitem__, ('432', slice(None)), 4.)
assert_raises(ValueError, ds.__setitem__, (slice(None), '432'), 4.)
# deleting items
del ds['A']
ok_('A' not in ds)
assert_raises(KeyError, ds.__getitem__, 'A')
del ds['B', 'rm']
ok_('B' not in ds and 'rm' not in ds)
def test_boxplot():
"Test plot.Boxplot"
ds = datasets.get_uv()
plot.Boxplot('fltvar', 'A%B', match='rm', ds=ds, show=False)
plt.close('all')
# many pairwise significances
ds['fltvar'][ds.eval("A%B==('a1','b1')")] += 1
ds['fltvar'][ds.eval("A%B==('a2','b2')")] -= 1
ds['C'] = Factor('qw', repeat=10, tile=4)
plot.Boxplot('fltvar', 'A%B%C', ds=ds, show=False)
def test_dataset_indexing():
"""Test Dataset indexing"""
ds = datasets.get_uv()
# indexing values
eq_(ds['A', 1], ds['A'][1])
eq_(ds[1, 'A'], ds['A'][1])
# indexing variables
assert_dataobj_equal(ds[:, 'A'], ds['A'])
assert_dataobj_equal(ds['A', :], ds['A'])
assert_dataobj_equal(ds[:10, 'A'], ds['A'][:10])
assert_dataobj_equal(ds['A', :10], ds['A'][:10])
# new Dataset through indexing
ds2 = Dataset()
ds2['A'] = ds['A']
assert_dataset_equal(ds[('A',)], ds2)
ds2['B'] = ds['B']
assert_dataset_equal(ds['A', 'B'], ds2)
assert_dataset_equal(ds[('A', 'B'), :10], ds2[:10])
assert_dataset_equal(ds[:10, ('A', 'B')], ds2[:10])
# assigning value
ds[2, 'A'] = 'hello'
eq_(ds[2, 'A'], 'hello')
ds['A', 2] = 'not_hello'
eq_(ds[2, 'A'], 'not_hello')
# assigning new factor
ds['C', :] = 'c'
ok_(np.all(ds.eval("C == 'c'")))
# assigning new Var
ds['D1', :] = 5.
ds[:, 'D2'] = 5.
assert_array_equal(ds['D1'], 5)
assert_array_equal(ds['D2'], 5)
# test illegal names
f = Factor('aaabbb')
assert_raises(ValueError, ds.__setitem__, '%dsa', f)
assert_raises(ValueError, ds.__setitem__, '432', f)
assert_raises(ValueError, ds.__setitem__, ('%dsa', slice(None)), 'value')
assert_raises(ValueError, ds.__setitem__, (slice(None), '%dsa'), 'value')
assert_raises(ValueError, ds.__setitem__, ('432', slice(None)), 4.)
assert_raises(ValueError, ds.__setitem__, (slice(None), '432'), 4.)
# deleting items
del ds['A']
ok_('A' not in ds)
assert_raises(KeyError, ds.__getitem__, 'A')
del ds['B', 'rm']
ok_('B' not in ds and 'rm' not in ds)
def test_boxplot():
"Test plot.uv.boxplot"
plot.configure_backend(False, False)
ds = datasets.get_uv()
plot.uv.Boxplot('fltvar', 'A%B', match='rm', ds=ds)
plt.close('all')
# many pairwise significances
ds['fltvar'][ds.eval("A%B==('a1','b1')")] += 1
ds['fltvar'][ds.eval("A%B==('a2','b2')")] -= 1
ds['C'] = Factor('qw', rep=10, tile=4)
plot.uv.Boxplot('fltvar', 'A%B%C', ds=ds)
def test_wilcoxon():
ds = datasets.get_uv()
ds_agg = ds.aggregate('A % rm', drop_bad=True)
n = ds_agg.n_cases // 2
w, p = scipy.stats.wilcoxon(ds_agg[:n, 'fltvar'].x,
ds_agg[n:, 'fltvar'].x,
alternative='two-sided')
res = test.WilcoxonSignedRank('fltvar', 'A', 'a1', 'a2', 'rm', ds=ds)
assert res.w == w
assert res.p == p
def test_boxplot():
"Test plot.uv.boxplot"
plot.configure_backend(False, False)
ds = datasets.get_uv()
plot.uv.Boxplot('fltvar', 'A%B', match='rm', ds=ds)
plt.close('all')
# many pairwise significances
ds['fltvar'][ds.eval("A%B==('a1','b1')")] += 1
ds['fltvar'][ds.eval("A%B==('a2','b2')")] -= 1
ds['C'] = Factor('qw', rep=10, tile=4)
plot.uv.Boxplot('fltvar', 'A%B%C', ds=ds)
def test_anova():
"Test ANOVA"
from rpy2.robjects import r
r_require('car')
r_require('ez')
ds = datasets.get_uv(nrm=True)
ds.to_r('ds')
# fixed effects
aov = test.ANOVA('fltvar', 'A*B', ds=ds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A*B', ds)
fnds = run_as_ndanova('fltvar', 'A*B', ds)
r_res = r("Anova(lm(fltvar ~ A * B, ds, type=2))")
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'Anova')
# random effect
aov = test.ANOVA('fltvar', 'A*B*rm', ds=ds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'A*B*rm', ds)
fnds = run_as_ndanova('fltvar', 'A*B*rm', ds)
r('test.aov <- aov(fltvar ~ A * B + Error(rm / (A * B)), ds)')
print(r('test.summary <- summary(test.aov)'))
r_res = r['test.summary'][1:]
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'rmaov')
# nested random effect
aov = test.ANOVA('fltvar', 'B + A%B + A * nrm(B)', ds=ds)
print(aov)
fs = run_on_lm_fitter('fltvar', 'B + A%B + A * nrm(B)', ds)
fnds = run_as_ndanova('fltvar', 'B + A%B + A * nrm(B)', ds)
r_res = r('ezANOVA(ds, fltvar, nrm, A, between=B)')
assert_f_tests_equal(aov.f_tests, r_res, fs, fnds, 'ez')
# not fully specified model with random effects
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*rm', ds=ds)
# empty cells
dss = ds.sub("A%B != ('a2', 'b2')")
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*B', ds=dss)
assert_raises(NotImplementedError,
run_on_lm_fitter,
'fltvar',
'A*B',
ds=dss)
dss = ds.sub("A%B != ('a1', 'b1')")
assert_raises(NotImplementedError, test.anova, 'fltvar', 'A*B', ds=dss)
assert_raises(NotImplementedError,
run_on_lm_fitter,
'fltvar',
'A*B',
ds=dss)
def test_tsv_io():
"""Test tsv I/O"""
tempdir = TempDir()
names = ['A', 'B', 'rm', 'intvar', 'fltvar', 'fltvar2', 'index']
ds = datasets.get_uv()
ds['fltvar'][5:10] = np.nan
ds[:4, 'rm'] = ''
# save and load
dst = Path(tempdir) / 'ds.txt'
ds.save_txt(dst)
ds1 = load.tsv(dst, random='rm')
assert_dataset_equal(ds1, ds, decimal=10)
ds1 = load.tsv(dst, skiprows=1, names=names, random='rm')
assert_dataset_equal(ds1, ds, decimal=10)
# delimiter
for delimiter in [' ', ',']:
ds.save_txt(dst, delimiter=delimiter)
ds1 = load.tsv(dst, delimiter=delimiter, random='rm')
assert_dataset_equal(ds1, ds, decimal=10)
# guess data types with missing
intvar2 = ds['intvar'].as_factor()
intvar2[10:] = ''
ds_intvar = Dataset((intvar2, ))
ds_intvar.save_txt(dst)
ds_intvar1 = load.tsv(dst, empty='nan')
assert_dataobj_equal(ds_intvar1['intvar', :10], ds['intvar', :10])
assert_array_equal(ds_intvar1['intvar', 10:], np.nan)
# str with space
ds[:5, 'A'] = 'a 1'
ds.save_txt(dst)
ds1 = load.tsv(dst, random='rm')
assert_dataset_equal(ds1, ds, decimal=10)
ds.save_txt(dst, delimiter=' ')
ds1 = load.tsv(dst, delimiter=' ', random='rm')
assert_dataset_equal(ds1, ds, decimal=10)
# Fixed column width
path = file_path('fox-prestige')
ds = load.tsv(path, delimiter=' ', skipinitialspace=True)
assert ds[1] == {
'id': 'GENERAL.MANAGERS',
'education': 12.26,
'income': 25879,
'women': 4.02,
'prestige': 69.1,
'census': 1130,
'type': 'prof'
}
def test_interaction():
"Test Interaction"
ds = datasets.get_uv()
A = ds['A']
B = ds['B']
i = A % B
# eq for sequence
assert_array_equal(i == A % B, True)
assert_array_equal(i == B % A, False)
assert_array_equal(i == A, False)
assert_array_equal(i == ds['fltvar'], False)
assert_array_equal(ds.eval("A%B") == Factor(ds['A']) % B, True)
# eq for element
for a, b in product(A.cells, B.cells):
assert_array_equal(i == (a, b), np.logical_and(A == a, B == b))
def test_barplot():
"Test plot.Barplot"
ds = datasets.get_uv()
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, show=False)
plot.Barplot('fltvar', 'A%B', match='rm', ds=ds, pool_error=False,
show=False)
plot.Barplot('fltvar', 'A%B', match='rm', test=0, ds=ds, show=False)
# Fixed top
p = plot.Barplot('fltvar', 'A%B', ds=ds, top=2, test_markers=False,
show=False)
ax = p._axes[0]
eq_(ax.get_ylim()[1], 2)
plt.close('all')
def test_interaction():
"Test Interaction"
ds = datasets.get_uv()
A = ds['A']
B = ds['B']
i = A % B
# eq for sequence
assert_array_equal(i == A % B, True)
assert_array_equal(i == B % A, False)
assert_array_equal(i == A, False)
assert_array_equal(i == ds['fltvar'], False)
assert_array_equal(ds.eval("A%B") == Factor(ds['A']) % B, True)
# eq for element
for a, b in product(A.cells, B.cells):
assert_array_equal(i == (a, b), np.logical_and(A == a, B == b))
def test_mann_whitney():
ds = datasets.get_uv()
ds_agg = ds.aggregate('A % rm', drop_bad=True)
n = ds_agg.n_cases // 2
a, b = ds_agg[:n, 'fltvar'], ds_agg[n:, 'fltvar']
u, p = scipy.stats.mannwhitneyu(a.x, b.x, alternative='two-sided')
res = test.MannWhitneyU('fltvar', 'A', 'a1', 'a2', 'rm', ds=ds)
assert res.u == u
assert res.p == p
res = test.MannWhitneyU(a, b)
assert res.u == u
assert res.p == p
def test_nested_effects():
"""Test nested effects"""
ds = datasets.get_uv(nrm=True)
nested = ds.eval("nrm(B)")
eq_(nested.cells, ds['nrm'].cells)
# interaction
i = ds.eval("A % nrm(B)")
expected_cells = tuple((case['A'], case['nrm']) for case in ds.itercases())
eq_(i.cells, expected_cells)
assert_has_no_empty_cells(ds.eval('A * B + nrm(B) + A % nrm(B)'))
i = ds.eval("nrm(B) % A")
expected_cells = sorted((case['nrm'], case['A']) for case in ds.itercases())
eq_(i.cells, tuple(expected_cells))
def test_repmeas():
"Test table.repmeas (repeated measures table)"
ds = datasets.get_uv()
print(table.repmeas('fltvar', 'A', 'rm', ds=ds))
print(table.repmeas('fltvar', 'A%B', 'rm', ds=ds))
print(table.repmeas('fltvar', 'A', 'B%rm', ds=ds))
# with int model
ds['Bv'] = ds['B'].as_var({'b1': 1, 'b2': 2})
print(table.repmeas('fltvar', 'A', 'Bv%rm', ds=ds))
# test naturalization of cellnames
ds['ANum'] = Factor(ds['A'], labels={'a1': '1', 'a2': '2'})
print(table.repmeas('fltvar', 'ANum', 'rm', ds=ds))
# with empty cell name
ds['A'].update_labels({'a1': ''})
print(table.repmeas('fltvar', 'A', 'rm', ds=ds))
def test_io_txt():
"Test Dataset io as text"
ds = datasets.get_uv()
# Var that has integer values as float
ds['intflt'] = ds.eval('intvar * 1.')
ds['intflt'].name = 'intflt'
# io test
tempdir = tempfile.mkdtemp()
try:
dest = os.path.join(tempdir, 'test.txt')
ds.save_txt(dest)
ds2 = load.tsv(dest)
finally:
shutil.rmtree(tempdir)
assert_dataset_equal(ds, ds2, decimal=6)
def test_io_txt():
"Test Dataset io as text"
ds = datasets.get_uv()
# Var that has integer values as float
ds['intflt'] = ds.eval('intvar * 1.')
ds['intflt'].name = 'intflt'
# io test
tempdir = tempfile.mkdtemp()
try:
dest = os.path.join(tempdir, 'test.txt')
ds.save_txt(dest)
ds2 = load.tsv(dest)
finally:
shutil.rmtree(tempdir)
assert_dataset_equal(ds, ds2, decimal=6)
def test_align():
"Testing align() and align1() functions"
ds = datasets.get_uv()
# index the dataset
ds.index()
ds['aindex'] = ds.eval("A.enumerate_cells()")
# subset
idx4 = np.arange(0, ds.n_cases, 4)
idx4i = idx4[::-1]
ds2 = ds.sub(np.arange(0, ds.n_cases, 2))
# shuffle the whole dataset
shuffle_index = np.arange(ds.n_cases)
np.random.shuffle(shuffle_index)
ds_shuffled = ds[shuffle_index]
# align1: align Dataset to index
dsa = align1(ds2, idx4)
assert_array_equal(dsa['index'], idx4, "align1() failure")
dsa = align1(ds2, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
# d_idx as Var
dsa = align1(ds2[::2], idx4, idx4i)
assert_array_equal(dsa['index'], idx4i, "align1() failure")
assert_raises(ValueError, align1, ds2, idx4, idx4i)
# Factor index
assert_raises(ValueError, align1, ds, ds['rm', ::-1], 'rm')
fds = ds[:20]
dsa = align1(fds, fds['rm', ::-1], 'rm')
assert_array_equal(dsa['index'], np.arange(19, -1, -1), "align1 Factor")
# align two datasets
dsa1, dsa2 = align(ds, ds2)
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
dsa1, dsa2 = align(ds, ds2[::-1])
assert_array_equal(dsa1['index'], dsa2['index'], "align() failure")
dsa1, dsa2 = align(ds, ds_shuffled)
assert_dataset_equal(dsa1, dsa2)
# align using categorial
dsa1, dsa2 = align(ds, ds_shuffled, 'A % aindex')
assert_dataset_equal(dsa1, dsa2)
dsa1, dsa2 = align(ds, ds_shuffled, 'aindex % A')
assert_dataset_equal(dsa1, dsa2)
def test_report():
"Test fmtxt.Report class"
tempdir = tempfile.mkdtemp()
report = fmtxt.Report("Test Report")
section = report.add_section(u'unicode: \xe2 abc')
ds = datasets.get_uv()
p = plot.Barplot('fltvar', 'A', sub="B=='b1'", ds=ds)
image = p.image()
section.add_figure("test", image)
report.sign()
# report output
print report
dst = os.path.join(tempdir, 'report.html')
report.save_html(dst)
# clean up
shutil.rmtree(tempdir)
