def new_nexus_without_sites(nexus_obj, sites_to_remove):
"""
Returns a new NexusReader instance with the sites in
`sites_to_remove` removed.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param sites_to_remove: A list of site numbers
:type sites_to_remove: List
:return: A NexusWriter instance
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
# make new nexus
nexout = NexusWriter()
nexout.add_comment(
"Removed %d sites: %s" %
(len(sites_to_remove), ",".join(["%s" % s for s in sites_to_remove]))
)
new_sitepos = 0
for sitepos in range(nexus_obj.data.nchar):
if sitepos in sites_to_remove:
continue # skip!
for taxon, data in nexus_obj.data:
nexout.add(taxon, new_sitepos, data[sitepos])
new_sitepos += 1
return nexout
def test_regression_format_string_has_quoted_symbols(self):
"""Regression: Symbols in the format string should be quoted"""
nex = NexusWriter()
for char, b in data.items():
for taxon, value in b.items():
nex.add(taxon, char, value)
out = nex.make_nexus()
assert 'SYMBOLS="123456"' in out
def test_regression_format_string_has_datatype_first(self):
"""
Regression: Format string should contain 'datatype' as the first
element
"""
# SplitsTree complains otherwise.
nex = NexusWriter()
for char, b in data.items():
for taxon, value in b.items():
nex.add(taxon, char, value)
out = nex.make_nexus()
assert "FORMAT DATATYPE=STANDARD" in out
def binarise(nexus_obj, one_nexus_per_block=False, keep_zero=False):
"""
Returns a binary variant of the given `nexus_obj`.
If `one_nexus_per_block` then we return a list of NexusWriter instances.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param one_nexus_per_block: Whether to return a single NexusWriter, or a
list of NexusWriter's (one per character)
:type one_nexus_per_block: Boolean
:param keep_zero: A boolean flag denoting whether to
treat '0' as a missing state or not. The default
(False) is to ignore '0' as a trait absence.
Setting this to True will treat '0' as a unique
state.
:type keep_zero: Boolean
:return: A NexusWriter instance or a list of NexusWriter instances.
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
nexuslist = []
n = NexusWriter()
for i in sorted(nexus_obj.data.charlabels):
label = nexus_obj.data.charlabels[i] # character label
char = nexus_obj.data.characters[label] # character dict
recoding = _recode_to_binary(char, keep_zero) # recode
new_char_length = len(recoding[list(recoding.keys())[0]])
# loop over recoded data
for j in range(new_char_length):
for taxon, state in recoding.items():
# make new label
new_label = "%s_%d" % (str(label), j)
# add to nexus
n.add(taxon, new_label, state[j])
if one_nexus_per_block:
nexuslist.append(n)
n = NexusWriter()
if one_nexus_per_block:
return nexuslist
else:
return n
def test_mixed_type_characters():
n = NexusWriter()
n.add('taxon1', 'Character1', 'A')
n.add('taxon2', 'Character1', 'C')
n.add('taxon3', 'Character1', 'A')
with pytest.raises(AssertionError):
n.add('taxon1', 2, 1, check=True)
def multistatise(nexus_obj, charlabel=None):
"""
Returns a multistate variant of the given `nexus_obj`.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:return: A NexusReader instance
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
charlabel = charlabel or getattr(nexus_obj, 'short_filename', 1) or 1
states = {}
for taxon in nexus_obj.data.matrix:
states[taxon] = []
sequence = nexus_obj.data.matrix[taxon]
for site_idx, value in enumerate(sequence):
if site_idx > 26:
raise ValueError(
"Too many characters to handle! - run out of A-Z")
assert value == str(value), "%r is not a string" % value
if value == '1':
states[taxon].append(chr(65 + site_idx))
nexout = NexusWriter()
for taxon in states:
if not states[taxon]:
nexout.add(taxon, charlabel, '?')
else:
for s in states[taxon]:
nexout.add(taxon, charlabel, s)
return nexout._convert_to_reader()
def writer():
data = {
'char1': {
'French': 1,
'English': 2,
'Latin': 3
},
'char2': {
'French': 4,
'English': 5,
'Latin': 6
},
}
res = NexusWriter()
for char in data:
for taxon, value in data[char].items():
res.add(taxon, char, value)
return res
def shufflenexus(nexus_obj, resample=False):
"""
Shuffles the characters between each taxon to create a new nexus
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param resample: The number of characters to resample. If set to False,
then the number of characters will equal the number of characters
in the original data file.
:type resample: Integer
:return: A shuffled NexusReader instance
:raises AssertionError: if nexus_obj is not a nexus
:raises ValueError: if resample is not False or a positive Integer
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
if resample is False:
resample = nexus_obj.data.nchar
try:
resample = int(resample)
except ValueError:
raise ValueError('resample must be a positive integer or False!')
if resample < 1:
raise ValueError('resample must be a positive integer or False!')
newnexus = NexusWriter()
newnexus.add_comment(
"Randomised Nexus generated from %s" % nexus_obj.filename
)
for i in range(resample):
# pick existing character
character = randrange(0, nexus_obj.data.nchar)
chars = nexus_obj.data.characters[character]
site_values = [chars[taxon] for taxon in nexus_obj.data.taxa]
shuffle(site_values)
for taxon in nexus_obj.data.taxa:
newnexus.add(taxon, i, site_values.pop(0))
return newnexus
def new_nexus_without_sites(nexus_obj, sites_to_remove):
"""
Returns a new NexusReader instance with the sites in
`sites_to_remove` removed.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param sites_to_remove: A list of site numbers
:type sites_to_remove: List
:return: A NexusWriter instance
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
# make new nexus
nexout = NexusWriter()
nexout.add_comment(
"Removed %d sites: %s" %
(len(sites_to_remove), ",".join(["%s" % s for s in sites_to_remove])))
new_sitepos = 0
for sitepos in range(nexus_obj.data.nchar):
if sitepos in sites_to_remove:
continue # skip!
for taxon, data in nexus_obj.data:
nexout.add(taxon, new_sitepos, data[sitepos])
new_sitepos += 1
return nexout
def combine_nexuses(nexuslist):
"""
Combines a list of NexusReader instances into a single nexus
:param nexuslist: A list of NexusReader instances
:type nexuslist: List
:return: A NexusWriter instance
:raises TypeError: if nexuslist is not a list of NexusReader instances
:raises IOError: if unable to read an file in nexuslist
:raises NexusFormatException: if a nexus file does not have a `data` block
"""
out = NexusWriter()
charpos = 0
for nex_id, nex in enumerate(nexuslist, 1):
check_for_valid_NexusReader(nex, required_blocks=['data'])
if hasattr(nex, 'short_filename'):
nexus_label = os.path.splitext(nex.short_filename)[0]
elif hasattr(nex, 'label'):
nexus_label = nex.label
else:
nexus_label = str(nex_id)
out.add_comment(
"%d - %d: %s" %
(charpos, charpos + nex.data.nchar - 1, nexus_label)
)
for site_idx, site in enumerate(sorted(nex.data.characters), 0):
data = nex.data.characters.get(site)
charpos += 1
# work out character label
charlabel = nex.data.charlabels.get(site_idx, site_idx + 1)
label = '%s.%s' % (nexus_label, charlabel)
for taxon, value in data.items():
out.add(taxon, label, value)
return out
def binarise(nexus_obj, keep_zero=False):
"""
Returns a binary variant of the given `nexus_obj`.
If `one_nexus_per_block` then we return a list of NexusWriter instances.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param keep_zero: A boolean flag denoting whether to
treat '0' as a missing state or not. The default
(False) is to ignore '0' as a trait absence.
Setting this to True will treat '0' as a unique
state.
:type keep_zero: Boolean
:return: A NexusWriter instance or a list of NexusWriter instances.
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
n = NexusWriter()
for i in sorted(nexus_obj.data.charlabels):
label = nexus_obj.data.charlabels[i] # character label
char = nexus_obj.data.characters[label] # character dict
recoding = _recode_to_binary(char, keep_zero) # recode
new_char_length = len(recoding[list(recoding.keys())[0]])
# loop over recoded data
for j in range(new_char_length):
for taxon, state in recoding.items():
# make new label
new_label = "%s_%d" % (str(label), j)
# add to nexus
n.add(taxon, new_label, state[j])
return n
def shufflenexus(nexus_obj, resample=False):
"""
Shuffles the characters between each taxon to create a new nexus
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:param resample: The number of characters to resample. If set to False,
then the number of characters will equal the number of characters
in the original data file.
:type resample: Integer
:return: A shuffled NexusReader instance
:raises AssertionError: if nexus_obj is not a nexus
:raises ValueError: if resample is not False or a positive Integer
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
if resample is False:
resample = nexus_obj.data.nchar
try:
resample = int(resample)
except ValueError:
raise ValueError('resample must be a positive integer or False!')
if resample < 1:
raise ValueError('resample must be a positive integer or False!')
newnexus = NexusWriter()
newnexus.add_comment("Randomised Nexus generated from %s" %
nexus_obj.filename)
for i in range(resample):
# pick existing character
character = randrange(0, nexus_obj.data.nchar)
chars = nexus_obj.data.characters[character]
site_values = [chars[taxon] for taxon in nexus_obj.data.taxa]
shuffle(site_values)
for taxon in nexus_obj.data.taxa:
newnexus.add(taxon, i, site_values.pop(0))
return newnexus
def multistatise(nexus_obj, charlabel=None):
"""
Returns a multistate variant of the given `nexus_obj`.
:param nexus_obj: A `NexusReader` instance
:type nexus_obj: NexusReader
:return: A NexusReader instance
:raises AssertionError: if nexus_obj is not a nexus
:raises NexusFormatException: if nexus_obj does not have a `data` block
"""
check_for_valid_NexusReader(nexus_obj, required_blocks=['data'])
if not charlabel:
charlabel = getattr(nexus_obj, 'short_filename', 1)
states = {}
for taxon in nexus_obj.data.matrix:
states[taxon] = []
sequence = nexus_obj.data.matrix[taxon]
for site_idx, value in enumerate(sequence):
if site_idx > 26:
raise ValueError(
"Too many characters to handle! - run out of A-Z"
)
assert value == str(value), "%r is not a string" % value
if value == '1':
states[taxon].append(chr(65 + site_idx))
nexout = NexusWriter()
for taxon in states:
if len(states[taxon]) == 0:
nexout.add(taxon, charlabel, '?')
else:
for s in states[taxon]:
nexout.add(taxon, charlabel, s)
return nexout._convert_to_reader()
def combine_nexuses(nexuslist):
"""
Combines a list of NexusReader instances into a single nexus
:param nexuslist: A list of NexusReader instances
:type nexuslist: List
:return: A NexusWriter instance
:raises TypeError: if nexuslist is not a list of NexusReader instances
:raises IOError: if unable to read an file in nexuslist
:raises NexusFormatException: if a nexus file does not have a `data` block
"""
out = NexusWriter()
charpos = 0
for nex_id, nex in enumerate(nexuslist, 1):
check_for_valid_NexusReader(nex, required_blocks=['data'])
if hasattr(nex, 'short_filename'):
nexus_label = os.path.splitext(nex.short_filename)[0]
else:
nexus_label = str(nex_id)
out.add_comment("%d - %d: %s" %
(charpos, charpos + nex.data.nchar - 1, nexus_label))
for site_idx, site in enumerate(sorted(nex.data.characters), 0):
data = nex.data.characters.get(site)
charpos += 1
# work out character label
charlabel = nex.data.charlabels.get(site_idx, site_idx + 1)
label = '%s.%s' % (nexus_label, charlabel)
for taxon, value in data.items():
out.add(taxon, label, value)
return out
class Test_NexusWriter(unittest.TestCase):
def setUp(self):
self.nex = NexusWriter()
for char in data:
for taxon, value in data[char].items():
self.nex.add(taxon, char, value)
def test_char_adding1(self):
"""Test Character Addition 1"""
assert self.nex.data['char1']['French'] == '1'
assert self.nex.data['char1']['English'] == '2'
assert self.nex.data['char1']['Latin'] == '3'
def test_char_adding2(self):
"""Test Character Addition 2"""
assert self.nex.data['char2']['French'] == '4'
assert self.nex.data['char2']['English'] == '5'
assert self.nex.data['char2']['Latin'] == '6'
def test_char_adding_integer(self):
"""Test Character Addition as integer"""
self.nex.add('French', 'char3', 9)
self.nex.add('English', 'char3', '9')
assert self.nex.data['char3']['French'] == '9'
assert self.nex.data['char3']['French'] == '9'
def test_characters(self):
assert 'char1' in self.nex.characters
assert 'char2' in self.nex.characters
def test_taxa(self):
assert 'French' in self.nex.taxa
assert 'English' in self.nex.taxa
assert 'Latin' in self.nex.taxa
def test_remove(self):
self.nex.remove("French", "char2")
assert 'French' not in self.nex.data['char2']
assert 'French' in self.nex.taxa
def test_remove_character(self):
self.nex.remove_character("char2")
assert len(self.nex.characters) == 1
assert 'char2' not in self.nex.data
def test_remove_taxon(self):
self.nex.remove_taxon("French")
assert 'French' not in self.nex.taxa
for char in self.nex.data:
assert 'French' not in self.nex.data[char]
n = self.nex.make_nexus(interleave=False)
assert re.search("DIMENSIONS NTAX=2 NCHAR=2;", n)
assert 'French' not in n
def test_nexus_noninterleave(self):
"""Test Nexus Generation - Non-Interleaved"""
n = self.nex.make_nexus(interleave=False)
assert re.search("#NEXUS", n)
assert re.search("BEGIN DATA;", n)
assert re.search("DIMENSIONS NTAX=3 NCHAR=2;", n)
assert re.search("MATRIX", n)
assert re.search("Latin\s+36", n)
assert re.search("French\s+14", n)
assert re.search("English\s+25", n)
assert re.search("FORMAT.*MISSING\=(.+?)", n).groups()[0] == '?'
assert re.search("FORMAT.*DATATYPE\=(\w+)\s", n).groups()[0] \
== 'STANDARD'
assert re.search('FORMAT.*SYMBOLS\="(\d+)";', n).groups()[0] \
== '123456'
def test_nexus_charblock(self):
"""Test Nexus Generation - with characters block"""
n = self.nex.make_nexus(charblock=True)
assert re.search("#NEXUS", n)
assert re.search("BEGIN DATA;", n)
assert re.search("DIMENSIONS NTAX=3 NCHAR=2;", n)
assert re.search("CHARSTATELABELS", n)
assert re.search("1 char1,", n)
assert re.search("2 char2", n)
assert re.search("MATRIX", n)
assert re.search("Latin\s+36", n)
assert re.search("French\s+14", n)
assert re.search("English\s+25", n)
assert re.search("FORMAT.*MISSING\=(.+?)", n).groups()[0] == '?'
assert re.search("FORMAT.*DATATYPE\=(\w+)\s", n).groups()[0] \
== 'STANDARD'
assert re.search('FORMAT.*SYMBOLS\="(\d+)";', n).groups()[0] \
== '123456'
def test_nexus_interleave(self):
"""Test Nexus Generation - Interleaved"""
n = self.nex.make_nexus(interleave=True)
assert re.search("#NEXUS", n)
assert re.search("BEGIN DATA;", n)
assert re.search("DIMENSIONS NTAX=3 NCHAR=2;", n)
assert re.search("MATRIX", n)
# char1
assert re.search("Latin\s+3", n)
assert re.search("French\s+1", n)
assert re.search("English\s+2", n)
# char2
assert re.search("Latin\s+6", n)
assert re.search("French\s+4", n)
assert re.search("English\s+5", n)
assert re.search("FORMAT.*MISSING\=(.+?)", n).groups()[0] == '?'
assert re.search("FORMAT.*DATATYPE\=(\w+)\s", n).groups()[0] == \
'STANDARD'
assert re.search("FORMAT.*(INTERLEAVE)", n).groups()[0] == \
'INTERLEAVE'
assert re.search('FORMAT.*SYMBOLS\="(\d+)";', n).groups()[0] == \
'123456'
def test_polymorphic_characters(self):
self.nex.add("French", "char1", 2)
self.assertEqual(self.nex.data['char1']['French'], "12")
n = self.nex.make_nexus(charblock=True)
assert re.search("DIMENSIONS NTAX=3 NCHAR=2;", n) # no change
assert re.search("French\s+\(12\)4", n)
def test_write_to_file(self):
tmp = NamedTemporaryFile(delete=False, suffix=".nex")
tmp.close()
self.nex.write_to_file(tmp.name)
assert os.path.isfile(tmp.name)
with open(tmp.name, 'r') as handle:
n = handle.read()
assert re.search("#NEXUS", n)
assert re.search("BEGIN DATA;", n)
assert re.search("DIMENSIONS NTAX=3 NCHAR=2;", n)
assert re.search("MATRIX", n)
assert re.search("Latin\s+36", n)
assert re.search("French\s+14", n)
assert re.search("English\s+25", n)
assert re.search("FORMAT.*MISSING\=(.+?)", n).groups()[0] == '?'
assert re.search("FORMAT.*DATATYPE\=(\w+)\s", n).groups()[0] \
== 'STANDARD'
assert re.search('FORMAT.*SYMBOLS\="(\d+)";', n).groups()[0] \
== '123456'
os.unlink(tmp.name) # cleanup
def test_write_as_table(self):
content = self.nex.write_as_table()
assert re.search("Latin\s+36", content)
assert re.search("French\s+14", content)
assert re.search("English\s+25", content)
assert len(content.split("\n")) == 3
def test_write_as_table_with_polymorphoc(self):
self.nex.add('French', 'char1', '2')
content = self.nex.write_as_table()
assert re.search("Latin\s+36", content)
assert re.search("French\s+\(12\)4", content)
assert re.search("English\s+25", content)
assert len(content.split("\n")) == 3
def setUp(self):
self.nex = NexusWriter()
for char in data:
for taxon, value in data[char].items():
self.nex.add(taxon, char, value)
def test_invalid():
w = NexusWriter()
with pytest.raises(ValueError):
w.write()
def test_write_with_no_data_but_trees():
nex = NexusWriter()
nex.trees.append('tree tree1 = (French,(English,Latin));')
content = nex.write()
assert re.search(r"BEGIN TREES", content)
assert re.search(r"tree tree1 = \(French,\(English,Latin\)\);", content)