def test_2(path, tmpdir, regtest):
# same folder as table for peakmaps:
from emzed.core.data_types import Table
pm = emzed.io.loadPeakMap(path("data/SHORT_MS2_FILE.mzData"))
t = emzed.utils.toTable("id", (1, 2, 3), type_=int)
t.addColumn("peakmap", pm, type_=object)
p2 = tmpdir.join("with_comp.table").strpath
t.store(p2, True, True, ".")
# chekc if pm proxy is next ot
file_names = [p.basename for p in tmpdir.listdir()]
print(file_names, file=regtest)
# force loading of data:.
print(len(t.peakmap.values[0]), file=regtest)
s1 = os.stat(p2).st_size
assert s1 < 1000
t = Table.load(p2)
print(len(t.peakmap.values[0]), file=regtest)
subfolder = tmpdir.join("subfolder")
os.makedirs(subfolder.strpath)
for p in file_names:
shutil.move(tmpdir.join(p).strpath, subfolder.strpath)
t2 = Table.load(subfolder.join("with_comp.table").strpath)
print(t2, file=regtest)
print(len(t2.peakmap.values[0]), file=regtest)
print(t2, file=regtest)
def test_3(path, tmpdir, regtest):
# relative path for peakmaps:
from emzed.core.data_types import Table
pm = emzed.io.loadPeakMap(path("data/SHORT_MS2_FILE.mzData"))
t = emzed.utils.toTable("id", (1, 2, 3), type_=int)
t.addColumn("peakmap", pm, type_=object)
folder = tmpdir.join("subfolder")
folder.mkdir()
p2 = folder.join("with_comp.table").strpath
t.store(p2, True, True, "..")
for p in tmpdir.listdir():
# full path varies because the folder is a tmp dir:
print(p.basename, file=regtest)
t2 = Table.load(p2)
# this triggers loading:
t2.peakmap[0].chromatogram(0, 10, 0, 10)
t2.peakmap[0]._path == p.strpath
# now we move and check if proxy still works:
moved = folder.join("..").join("moved")
shutil.move(folder.strpath, moved.strpath)
t2 = Table.load(moved.join("with_comp.table").strpath)
t2.peakmap[0].chromatogram(0, 10, 0, 10)
def testSplitBy(self):
t = toTable("a", [1, 1, 3, 4])
t.addColumn("b", [1, 1, 3, 3])
t.addColumn("c", [1, 2, 1, 4])
t._print()
subts = t.splitBy("a")
assert len(subts) == 3
res = Table.mergeTables(subts)
assert len(res) == len(t)
subts[0]._print()
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == ["a", "b", "c"]
subts = t.splitBy("a", "c")
assert len(subts) == 4
res = Table.mergeTables(subts)
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == ["a", "b", "c"]
def testSplitBy(self):
t = toTable("a", [1,1,3,4])
t.addColumn("b", [1,1,3,3])
t.addColumn("c", [1,2,1,4])
t._print()
subts = t.splitBy("a")
assert len(subts) == 3
res = Table.mergeTables(subts)
assert len(res) == len(t)
subts[0]._print()
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == [ "a", "b", "c"]
subts = t.splitBy("a", "c")
assert len(subts) == 4
res = Table.mergeTables(subts)
assert res.a.values == t.a.values
assert res.b.values == t.b.values
assert res.c.values == t.c.values
# check if input tables are not altered
for subt in subts:
assert subt.getColNames() == [ "a", "b", "c"]
def testSomePredicates(self):
#build table
names="int long float str object array".split()
types = [int, long, float, str, object, np.ndarray,]
formats = [ "%3d", "%d", "%.3f", "%s", "%r", "'array%r' % (o.shape,)" ]
row1 = [ 1, 12323L, 1.0, "hi", { 1: 1 }, np.array((1,2,3)) ]
row2 = [ 2, 22323L, 2.0, "hi2", [2,3,], np.array(((2,3,4),(1,2,3))) ]
row3 = [ 3, 32323L, 3.0, "hi3", (3,) , np.array(((3,3,4,5),(1,2,3,4))) ]
rows = [row1, row2, row3]
t=Table(names, types, formats, rows, "testtabelle", meta=dict(why=42))
expr = (t.int + t.float).inRange(-1, 2)
tn = t.filter(expr)
assert len(tn) == 1
assert tn.getValue(tn.rows[0], "int") == 1
tn = t.filter((t.float+t.int).inRange(-1, 2))
assert len(tn) == 1
assert tn.getValue(tn.rows[0], "int") == 1
tn = t.filter(t.float.approxEqual(1.0, t.int/10))
tn._print()
assert len(tn) == 1, len(tn)
assert tn.getValue(tn.rows[0], "int") == 1
def testDynamicColumnAttributes(self):
t = Table(["a", "b", "c"], [str, int, int], ["%s", "%d", "%d"], [])
t.a
t.b
t.c
assert len(t.a.values) == 0
assert len(t.b.values) == 0
assert len(t.c.values) == 0
t.renameColumns(dict(a="aa"))
assert "a" not in t.getColNames()
assert "aa" in t.getColNames()
t.aa
try:
t.a
raise Exception("t.a should be deteted")
except:
pass
col = pickle.loads(pickle.dumps(t.aa))
assert len(col.values) == 0
t.dropColumns("aa")
assert "aa" not in t.getColNames()
try:
t.aa
raise Exception("t.aa should be deteted")
except:
pass
def testToOpenMSFeatureMap():
t = Table("mz rt".split(), [float, float], 2 * ["%.6f"])
fm = t.toOpenMSFeatureMap()
assert fm.size() == 0
t.addRow([1.0, 2.0])
fm = t.toOpenMSFeatureMap()
assert fm.size() == 1
f = fm[0]
assert f.getMZ() == 1.0 # == ok, as no digits after decimal point
assert f.getRT() == 2.0 # dito
def test_removePostfixes(regtest):
t = Table._create(["abb__0", "bcb__0"], [str] * 2, ["%s"] * 2)
assert t.getColNames() == ["abb__0", "bcb__0"]
t.removePostfixes()
assert t.getColNames() == ["abb", "bcb"]
t.removePostfixes("bb", "cb")
assert t.getColNames() == ["a", "b"]
t = Table._create(["abb__0", "bcb__0", "abb"], [str] * 3, ["%s"] * 3)
with pytest.raises(Exception) as e:
t.removePostfixes()
print(e.value, file=regtest)
def test_removePostfixes(regtest):
t = Table._create(["abb__0", "bcb__0"], [str] * 2, ["%s"] * 2)
assert t.getColNames() == ["abb__0", "bcb__0"]
t.removePostfixes()
assert t.getColNames() == ["abb", "bcb"]
t.removePostfixes("bb", "cb")
assert t.getColNames() == ["a", "b"]
t = Table._create(["abb__0", "bcb__0", "abb"], [str] * 3, ["%s"] * 3)
with pytest.raises(Exception) as e:
t.removePostfixes()
print(e.value, file=regtest)
def testDynamicColumnAttributes(self):
t = Table(["a", "b", "c"], [str, int, int], ["%s", "%d", "%d"],[])
t.a
t.b
t.c
assert len(t.a.values) == 0
assert len(t.b.values) == 0
assert len(t.c.values) == 0
t.renameColumns(dict(a="aa"))
assert "a" not in t.getColNames()
assert "aa" in t.getColNames()
t.aa
try:
t.a
raise Exception("t.a should be deteted")
except:
pass
col = pickle.loads(pickle.dumps(t.aa))
assert len(col.values) == 0
t.dropColumns("aa")
assert "aa" not in t.getColNames()
try:
t.aa
raise Exception("t.aa should be deteted")
except:
pass
def setupTable():
names = "int long float str object array".split()
types = [
int,
long,
float,
str,
object,
np.ndarray,
]
formats = ["%3d", "%d", "%.3f", "%s", "%r", "'array(%r)' % o.shape"]
row1 = [1, 12323L, 1.0, "hi", {1: 1}, np.array((1, 2, 3))]
row2 = [2, 22323L, 2.0, "hi2", [
2,
3,
], np.array(((2, 3, 4), (1, 2, 3)))]
row3 = [
3, 32323L, 3.0, "hi3", (3, ),
np.array(((3, 3, 4, 5), (1, 2, 3, 4)))
]
rows = [row1, row2, row3]
t = Table(names, types, formats, rows, "testtabelle", meta=dict(why=42))
t = t.extractColumns("int", "float", "str")
t.addEnumeration()
t._name = "t"
t._print()
return t
def testDoubleColumnames(self):
ex = None
try:
colnames = ["col0", "col0", "col1", "col1", "col2"]
Table(colnames, [int] * 5, [""] * 5)
except Exception, e:
ex = e.message
def testMerge(self):
t1 = toTable("a", [1])
t1.addColumn("b", [2])
t1.addColumn("c", [3])
t2 = toTable("a", [1, 2])
t2.addColumn("c", [1, 3])
t2.addColumn("d", [1, 4])
tn = Table.mergeTables([t1, t2])
assert tn.a.values == (
1,
1,
2,
)
assert tn.b.values == (
2,
None,
None,
)
assert tn.c.values == (
3,
1,
3,
)
assert tn.d.values == (
None,
1,
4,
)
# check if input tables are not altered
assert t1.getColNames() == ["a", "b", "c"]
assert t2.getColNames() == ["a", "c", "d"]
def testSomePredicates(self):
#build table
names = "int long float str object array".split()
types = [
int,
long,
float,
str,
object,
np.ndarray,
]
formats = ["%3d", "%d", "%.3f", "%s", "%r", "'array%r' % (o.shape,)"]
row1 = [1, 12323L, 1.0, "hi", {1: 1}, np.array((1, 2, 3))]
row2 = [
2, 22323L, 2.0, "hi2", [
2,
3,
],
np.array(((2, 3, 4), (1, 2, 3)))
]
row3 = [
3, 32323L, 3.0, "hi3", (3, ),
np.array(((3, 3, 4, 5), (1, 2, 3, 4)))
]
rows = [row1, row2, row3]
t = Table(names,
types,
formats,
rows,
"testtabelle",
meta=dict(why=42))
expr = (t.int + t.float).inRange(-1, 2)
tn = t.filter(expr)
assert len(tn) == 1
assert tn.getValue(tn.rows[0], "int") == 1
tn = t.filter((t.float + t.int).inRange(-1, 2))
assert len(tn) == 1
assert tn.getValue(tn.rows[0], "int") == 1
tn = t.filter(t.float.approxEqual(1.0, t.int / 10))
tn._print()
assert len(tn) == 1, len(tn)
assert tn.getValue(tn.rows[0], "int") == 1
def testSupportedPostfixes():
names = "mz mzmin mzmax mz0 mzmin0 mzmax0 mz1 mzmax1 mzmin__0 mzmax__0 mz__0 "\
"mzmax3 mz4 mzmin4".split()
t = Table._create(names, [float] * len(names), [None] * len(names))
assert len(t.supportedPostfixes(["mz"])) == len(names)
assert t.supportedPostfixes(["mz", "mzmin"]) == ["", "0", "4", "__0"]
assert t.supportedPostfixes(["mz", "mzmin", "mzmax"]) == ["", "0", "__0"]
def testSupportedPostfixes():
names = "mz mzmin mzmax mz0 mzmin0 mzmax0 mz1 mzmax1 mzmin__0 mzmax__0 mz__0 "\
"mzmax3 mz4 mzmin4".split()
t = Table._create(names, [float] * len(names), [])
assert len(t.supportedPostfixes(["mz"])) == len(names)
assert t.supportedPostfixes(["mz", "mzmin"]) == ["", "0", "4", "__0"]
assert t.supportedPostfixes(["mz", "mzmin", "mzmax"]) == ["", "0", "__0"]
def test_pandas(regtest):
import pandas
from emzed.core.data_types import Table
data = dict(a=[1, 2, 3], b=[1.0, 2.0, None], c=["a", "b", "c"], d=[1.0, 2.0, 3.0],
e=[None, None, None])
df = pandas.DataFrame(data, columns=sorted(data.keys()))
t = Table.from_pandas(df, formats={"b": "%.2f", float: "%.1f", object: "%s", None: "%s"})
t.print_()
t.print_(out=regtest)
def test_removePostfixes():
t = Table._create(["abb__0", "bcb__0"], [str] * 2, ["%s"] * 2)
assert t.getColNames() == ["abb__0", "bcb__0"]
t.removePostfixes()
assert t.getColNames() == ["abb", "bcb"]
t.removePostfixes("bb", "cb")
assert t.getColNames() == ["a", "b"]
try:
t.print_()
t.removePostfixes("a", "b")
t.print_()
except:
pass
else:
assert False, "expected exception"
def testToOpenMSFeatureMap():
t = Table("mz rt".split(), [float, float], 2 * ["%.6f"])
fm = t.toOpenMSFeatureMap()
assert fm.size() == 0
t.addRow([1.0, 2.0])
fm = t.toOpenMSFeatureMap()
assert fm.size() == 1
f = fm[0]
assert f.getMZ() == 1.0 # == ok, as no digits after decimal point
assert f.getRT() == 2.0 # dito
def setupTable():
names="int long float str object array".split()
types = [int, long, float, str, object, np.ndarray,]
formats = [ "%3d", "%d", "%.3f", "%s", "%r", "'array(%r)' % o.shape" ]
row1 = [ 1, 12323L, 1.0, "hi", { 1: 1 }, np.array((1,2,3)) ]
row2 = [ 2, 22323L, 2.0, "hi2", [2,3,], np.array(((2,3,4),(1,2,3))) ]
row3 = [ 3, 32323L, 3.0, "hi3", (3,) , np.array(((3,3,4,5),(1,2,3,4))) ]
rows = [row1, row2, row3]
t=Table(names, types, formats, rows, "testtabelle", meta=dict(why=42))
t = t.extractColumns("int", "float", "str")
t.addEnumeration()
t._name = "t"
t._print()
return t
def testMixedRows():
names = """rt rtmin rtmax rt__1 rtmin__1 rtmax__1
mz mzmin mzmax mz__1 mzmin__1 mzmax__1
rt__0 rtmin__0 rtmax__0
mz__0 mzmin__0 mzmax__0
rt__2 rtmin__2""".split()
names = [n.strip() for n in names]
tab = Table._create(names, [float]*len(names), ["%f"] * len(names))
model = TableModel(tab, None)
#assert model.postfixes == [ "", "__0", "__1", "__2"], model.postfixes
assert model.table.supportedPostfixes(["rtmin", "rt"]) == ["", "__0", "__1", "__2"]
assert model.table.supportedPostfixes(["mz", "rt", "mzmin"]) == ["", "__0", "__1"]
assert model.table.supportedPostfixes(["mz", "rt", "x"]) == []
def loadExcel(path=None, sheetname=0, types=None, formats=None):
"""`sheetname` is either an intger or string for indicating the sheet which will be extracted
from the .xls or .xlsx file. The index 0 refers to the first sheet.
`types` is either None or a dictionary mapping column names to their types.
`formats` is either None or a dictionary mapping column names to formats.
"""
path = _prepare_path(path, extensions=["xls", "xlsx"])
if path is None:
return None
from emzed.core.data_types import Table
import pandas
# sheetname is reuqired for pandas < 0.14.0, later versions have default 0
df = pandas.read_excel(path, sheetname=sheetname)
return Table.from_pandas(df, types=types, formats=formats)
def test_pandas(regtest):
import pandas
from emzed.core.data_types import Table
data = dict(a=[1, 2, 3],
b=[1.0, 2.0, None],
c=["a", "b", "c"],
d=[1.0, 2.0, 3.0],
e=[None, None, None])
df = pandas.DataFrame(data, columns=sorted(data.keys()))
t = Table.from_pandas(df,
formats={
"b": "%.2f",
float: "%.1f",
object: "%s",
None: "%s"
})
t.print_()
t.print_(out=regtest)
def testMerge(self):
t1 = toTable("a", [1])
t1.addColumn("b", [2])
t1.addColumn("c", [3])
t2 = toTable("a", [1,2])
t2.addColumn("c", [1,3])
t2.addColumn("d", [1,4])
tn = Table.mergeTables([t1, t2])
assert tn.a.values == (1, 1, 2,)
assert tn.b.values == (2, None, None,)
assert tn.c.values == (3, 1, 3,)
assert tn.d.values == (None, 1, 4,)
# check if input tables are not altered
assert t1.getColNames() == [ "a", "b", "c"]
assert t2.getColNames() == [ "a", "c", "d"]
def test_pandas():
import pandas
import cStringIO
from emzed.core.data_types import Table
data = dict(a=[1, 2, 3], b=[1.0, 2.0, None], c=["a", "b", "c"], d=[1.0, 2.0, 3.0],
e=[None, None, None])
df = pandas.DataFrame(data, columns=sorted(data.keys()))
t = Table.from_pandas(df, formats={"b": "%.2f", float: "%.1f", object: "%s", None: "%s"})
out = cStringIO.StringIO()
t.print_()
t.print_(out=out)
lines = out.getvalue().split("\n")
assert len(lines) == 7
assert lines[0].strip() == "a b c d e"
assert lines[1].strip() == "int float str float None"
assert lines[2].strip() == "------ ------ ------ ------ ------"
assert lines[3].strip() == "1 1.00 a 1.0 -"
assert lines[4].strip() == "2 2.00 b 2.0 -"
assert lines[5].strip() == "3 - c 3.0 -"
def testIllegalRows():
with pytest.raises(Exception):
Table(["a", "b"], [float, float], ["%f", "%f"], [(1, 2)])
def testRunnerTable(self):
def run(t, colnames, rows):
t = t.copy()
ixs = set()
for i, name in enumerate(colnames):
ixs.add(t.getIndex(name))
assert not None in ixs
assert len(ixs) == len(colnames)
# test iteration
content = []
for row in t.rows:
for (cell, formatter) in zip(row, t.colFormatters):
content.append( formatter(cell))
break
# test formatting
assert content[0] == " 1"
assert content[1] == "12323"
assert content[2] == "1.000"
assert content[3] == "hi"
assert content[4] == repr({1:1})
assert content[5] in ("array((3,))", "array((3L,))"), content[5]
assert set(t.getVisibleCols()) == { 'int', 'long', 'float', 'str',
'object', 'array' }
expr = t.str.contains("hi")
tn = t.filter(expr)
assert len(tn) == 3
tn = t.filter(~ t.str.contains("hi"))
assert len(tn) == 0
tn = t.filter(t.str.contains("2"))
assert len(tn) == 1
# test requireColumn
for name in colnames:
t.requireColumn(name)
ex = None
try:
t.requireColumn("asdfkl?dsflkjaldfkja?sdlfjal?djf")
except Exception, e:
ex = e
assert ex is not None
# test other fields
assert t.meta["why"] == 42
assert t.title == "testtabelle"
t.sortBy("int", ascending=False)
# restrct cols
tn = t.extractColumns("int", "long")
assert len(tn.getColNames()) == 2, len(t.getColNames())
assert len(tn.getColTypes()) == 2
assert len(tn.getColFormats()) == 2
assert len(tn) == len(t)
assert tn.meta["why"] == 42
assert tn.title == "testtabelle"
tn.addEnumeration()
assert set(tn.getVisibleCols()) == { 'int', 'long', 'id' }
assert tn.getColNames()[0]=="id"
assert list(tn.id) == range(len(t))
tn.renameColumns(int='iii')
assert set(tn.getVisibleCols()) == { 'iii', 'long', 'id' }
tn.addColumn('x', 'hi', str, '%s')
assert set(tn.getVisibleCols()) == { 'iii', 'long', 'id', 'x' }
assert tn.getColNames()[-1]=="x"
assert list(tn.x) == ["hi"]*len(tn)
import os.path
def j(name):
return os.path.join(self.temp_output, name)
before = set(os.listdir(self.temp_output))
tn.storeCSV(j("x.csv"))
tnre = Table.loadCSV(j("x.csv"))
assert len(tnre) == len(tn)
assert tnre.getColNames() == tn.getColNames()
assert tnre.id.values == tn.id.values
assert tnre.iii.values == tn.iii.values
assert tnre.long.values == tn.long.values
assert tnre.x.values == tn.x.values
tn.storeCSV(j("x.csv"), onlyVisibleColumns=False)
after = set(os.listdir(self.temp_output))
# file written twice !
assert len(after-before) == 2
for n in after-before:
# maybe we have some x.csv.? from previous run of this
#function so we can not assume that we find x.csv and
#x.csv.1
assert re.match("x.csv(.\d+)?", n)
ex = None
with self.assertRaises(Exception):
# wrong file extension
tn.storeCSV(j("x.dat"))
# computed by exrpression
tn.print_()
tn.addColumn("computed", tn.long / (tn.iii + 1))
# computed by callback:
tn.addColumn("squared", lambda t,r,n: t.getValue(r, "iii")**2)
assert tn.getColumn("computed").values == (8080, 7441, 6161), tn.computed.values
assert tn.getColumn("squared").values == (9, 4, 1)
tn.replaceColumn("squared", tn.squared+1)
assert tn.getColumn("squared").values == (10, 5, 2)
assert len(tn.getColNames()) == 6
tx = tn.copy()
tx.dropColumns("squared", "computed")
assert tx.getColNames() == ["id", "iii", "long", "x"]
assert len(tx) == 3
tn.dropColumns("computed", "squared")
assert tn.getColNames() == ["id", "iii", "long", "x"]
assert len(tn) == 3
tn.dropColumns("id", "x")
t2 = tn.copy()
res = tn.leftJoin(t2, tn.iii == tn.long)
assert len(res) == len(t2)
res = tn.leftJoin(t2, tn.iii == tn.iii)
assert len(res) == len(t2)**2
res = tn.leftJoin(t2, (tn.iii == tn.iii) & (t2.long==32323))
assert len(res) == len(t2)
res = tn.join(t2, tn.iii == tn.long)
assert len(res) == 0
res = tn.join(t2, tn.long == tn.iii)
assert len(res) == 0
res = tn.join(t2, tn.iii == tn.iii)
assert len(res) == len(t2)**2, len(res)
res = tn.join(t2, (tn.iii == tn.iii) & (t2.long==32323))
assert len(res) == len(t2), len(res)
tx = tn.filter(tn.iii.isIn([1,4]))
tx._print()
assert len(tx) == 1
assert tx.iii.values == (1,)
tn.addColumn("li", [1,2,3])
assert len(tn) == 3
assert len(tn.getColNames()) == 3
assert "li" in tn.getColNames()
tn.addRow([1, 1, 1])
assert len(tn) == 4
ex = None
try:
tn.addRow([1,2,3,2])
except Exception, e:
ex = e
assert ex is not None
with self.assertRaises(Exception):
Table(["a"],[np.float32],["%f"],[[32.0]])
#build table
names="int long float str object array".split()
types = [int, long, float, str, object, np.ndarray,]
formats = [ "%3d", "%d", "%.3f", "%s", "%r", "'array(%r)' % (o.shape,)" ]
row1 = [ 1, 12323L, 1.0, "hi", { 1: 1 }, np.array((1,2,3)) ]
row2 = [ 2, 22323L, 2.0, "hi2", [2,3,], np.array(((2,3,4),(1,2,3))) ]
row3 = [ 3, 32323L, 3.0, "hi3", (3,) , np.array(((3,3,4,5),(1,2,3,4))) ]
rows = [row1, row2, row3]
t=Table(names, types, formats, rows, "testtabelle", meta=dict(why=42))
run(t, names, [row1, row2, row3])
# test pickle
dat = pickle.dumps(t)
t = pickle.loads(dat)
run(t, names, [row1, row2, row3])
def j(name):
import os.path
return os.path.join(self.temp_output, name)
t.store(j("test.table"))
try:
Table.storeTable(t, "temp_output/test.table")
def testIfThenElse(self):
t = Table(["a", "b", "c"], [str, int, int], ["%s", "%d", "%d"],[])
t.rows.append(["0", 1, 2])
t.rows.append([None, 2, 1])
t._print()
t.addColumn("x", (t.a.isNotNone()).thenElse(t.b, t.c))
assert t.getColNames()==["a", "b", "c", "x"]
print
t._print()
t.addColumn("y", (t.a.isNotNone()).thenElse("ok", "not ok"))
t._print()
assert t.y.values == ("ok", "not ok")
ex = None
try:
tn.addRow([1, 2, 3, 2])
except Exception, e:
ex = e
assert ex is not None
ex = None
try:
tn.addRow(["a", 1, 2])
except Exception, e:
ex = e
assert ex is not None
with self.assertRaises(Exception):
Table(["a"], [np.float32], ["%f"], [[32.0]])
#build table
names = "int long float str object array".split()
types = [
int,
long,
float,
str,
object,
np.ndarray,
]
formats = ["%3d", "%d", "%.3f", "%s", "%r", "'array(%r)' % (o.shape,)"]
row1 = [1, 12323L, 1.0, "hi", {1: 1}, np.array((1, 2, 3))]
row2 = [
def testRunnerTable(self):
def run(t, colnames, rows):
t = t.copy()
ixs = set()
for i, name in enumerate(colnames):
ixs.add(t.getIndex(name))
assert not None in ixs
assert len(ixs) == len(colnames)
# test iteration
content = []
for row in t.rows:
for (cell, formatter) in zip(row, t.colFormatters):
content.append(formatter(cell))
break
# test formatting
assert content[0] == " 1"
assert content[1] == "12323"
assert content[2] == "1.000"
assert content[3] == "hi"
assert content[4] == repr({1: 1})
assert content[5] in ("array((3,))", "array((3L,))"), content[5]
assert set(t.getVisibleCols()) == {
'int', 'long', 'float', 'str', 'object', 'array'
}
expr = t.str.contains("hi")
tn = t.filter(expr)
assert len(tn) == 3
tn = t.filter(~t.str.contains("hi"))
assert len(tn) == 0
tn = t.filter(t.str.contains("2"))
assert len(tn) == 1
# test requireColumn
for name in colnames:
t.requireColumn(name)
ex = None
try:
t.requireColumn("asdfkl?dsflkjaldfkja?sdlfjal?djf")
except Exception, e:
ex = e
assert ex is not None
# test other fields
assert t.meta["why"] == 42
assert t.title == "testtabelle"
t.sortBy("int", ascending=False)
# restrct cols
tn = t.extractColumns("int", "long")
assert len(tn.getColNames()) == 2, len(t.getColNames())
assert len(tn.getColTypes()) == 2
assert len(tn.getColFormats()) == 2
assert len(tn) == len(t)
assert tn.meta["why"] == 42
assert tn.title == "testtabelle"
tn.addEnumeration()
assert set(tn.getVisibleCols()) == {'int', 'long', 'id'}
assert tn.getColNames()[0] == "id"
assert list(tn.id) == range(len(t))
tn.renameColumns(int='iii')
assert set(tn.getVisibleCols()) == {'iii', 'long', 'id'}
tn.addColumn('x', 'hi', str, '%s')
assert set(tn.getVisibleCols()) == {'iii', 'long', 'id', 'x'}
assert tn.getColNames()[-1] == "x"
assert list(tn.x) == ["hi"] * len(tn)
import os.path
def j(name):
return os.path.join(self.temp_output, name)
before = set(os.listdir(self.temp_output))
tn.storeCSV(j("x.csv"))
tnre = Table.loadCSV(j("x.csv"))
assert len(tnre) == len(tn)
assert tnre.getColNames() == tn.getColNames()
assert tnre.id.values == tn.id.values
assert tnre.iii.values == tn.iii.values
assert tnre.long.values == tn.long.values
assert tnre.x.values == tn.x.values
with self.assertRaises(Exception):
# wrong file extension
tn.storeCSV(j("x.dat"))
# computed by exrpression
tn.print_()
tn.addColumn("computed", tn.long / (tn.iii + 1))
# computed by callback:
tn.addColumn("squared", lambda t, r, n: t.getValue(r, "iii")**2)
assert tn.getColumn("computed").values == (
8080, 7441, 6161), tn.computed.values
assert tn.getColumn("squared").values == (9, 4, 1)
tn.replaceColumn("squared", tn.squared + 1)
assert tn.getColumn("squared").values == (10, 5, 2)
assert len(tn.getColNames()) == 6
tx = tn.copy()
tx.dropColumns("squared", "computed")
assert tx.getColNames() == ["id", "iii", "long", "x"]
assert len(tx) == 3
tn.dropColumns("computed", "squared")
assert tn.getColNames() == ["id", "iii", "long", "x"]
assert len(tn) == 3
tn.dropColumns("id", "x")
t2 = tn.copy()
res = tn.leftJoin(t2, tn.iii == tn.long)
assert len(res) == len(t2)
res = tn.leftJoin(t2, tn.iii == tn.iii)
assert len(res) == len(t2)**2
res = tn.leftJoin(t2, (tn.iii == tn.iii) & (t2.long == 32323))
assert len(res) == len(t2)
res = tn.join(t2, tn.iii == tn.long)
assert len(res) == 0
res = tn.join(t2, tn.long == tn.iii)
assert len(res) == 0
res = tn.join(t2, tn.iii == tn.iii)
assert len(res) == len(t2)**2, len(res)
res = tn.join(t2, (tn.iii == tn.iii) & (t2.long == 32323))
assert len(res) == len(t2), len(res)
tx = tn.filter(tn.iii.isIn([1, 4]))
tx._print()
assert len(tx) == 1
assert tx.iii.values == (1, )
tn.addColumn("li", [1, 2, 3])
assert len(tn) == 3
assert len(tn.getColNames()) == 3
assert "li" in tn.getColNames()
tn.addRow([1, 1, 1])
assert len(tn) == 4
ex = None
try:
tn.addRow([1, 2, 3, 2])
except Exception, e:
ex = e
def testDetectionOfUnallowdColumnNames(self):
ex = None
try:
Table(["__init__"], [int], ["%d"])
except Exception, e:
ex = e.message
def testIfThenElse(self):
t = Table(["a", "b", "c"], [str, int, int], ["%s", "%d", "%d"], [])
t.rows.append(["0", 1, 2])
t.rows.append([None, 2, 1])
t._print()
t.addColumn("x", (t.a.isNotNone()).thenElse(t.b, t.c))
assert t.getColNames() == ["a", "b", "c", "x"]
print
t._print()
t.addColumn("y", (t.a.isNotNone()).thenElse("ok", "not ok"))
t._print()
assert t.y.values == ("ok", "not ok")
