def test_merge(self):
for merge_type in ("min", "max", "flat"):
ordered_bed_list = [
gqltools.load_file("../data/a.bed", "auto"),
gqltools.load_file("../data/b.bed", "auto"),
gqltools.load_file("../data/d.bed", "auto"),
]
mods = {}
R = gqltools.merge_beds(merge_type, ordered_bed_list, mods)
self.assertEqual(type(R), gqltypes.BED3)
was_error = False
try:
mods = {"distance": 10}
R = gqltools.merge_beds("flat", ordered_bed_list, mods)
except Exception:
was_error = True
self.assertTrue(was_error)
was_error = False
try:
mods = {"name": "OTHER"}
R = gqltools.merge_beds("flat", ordered_bed_list, mods)
except Exception:
was_error = True
self.assertTrue(was_error)
for func in ["MIN", "MAX", "SUM", "MEAN", "MEDIAN", "MODE", "ANITMODE", "COLLAPSE", "COUNT"]:
mods = {"score": func}
R = gqltools.merge_beds(merge_type, ordered_bed_list, mods)
self.assertEqual(type(R), gqltypes.BED6)
ordered_bed_list = [gqltools.load_file("../data/tiny*.bed", "auto")]
for func in ["MIN", "MAX", "SUM", "MEAN", "MEDIAN", "MODE", "ANITMODE", "COLLAPSE", "COUNT"]:
mods = {"score": func}
R = gqltools.merge_beds(merge_type, ordered_bed_list, mods)
self.assertEqual(type(R), gqltypes.BED6)
def eval_exp(exp, env):
# print exp # DEBUG
etype = exp[0]
# {{{ if etype == "identifier":
if etype == "identifier":
# print exp #Debug
# ('identifier', 'b')
vname = exp[1]
value = env_lookup(vname, env)
if value == None:
# print "ERROR: unbound variable " + vname
raise InterpException("Unbound variable " + vname, "identifier")
else:
return value
# }}}
# {{{ if etype == 'load':
if etype == "load":
# ('load',
# ('file', ('string', 'file0')),
# ('filetype', 'GENOME'))
file_path = eval_exp(exp[1], env)
# filetype_name = eval_exp(exp[2], env)
filetype_name = "auto"
# make sure the type is correct
return gqltools.load_file(file_path, filetype_name)
# }}}
# {{{ if etype == 'cast':
if etype == "cast":
# print exp
# ('cast', ('identifier', 'a'), ('filetype', 'BED6'), [])
ident = exp[1]
bedx = eval_exp(ident, env)
new_type = eval_exp(exp[2], env)
return gqltools.cast(bedx, new_type)
# }}}
# {{{ if etype == 'complement':
if etype == "complement":
# print exp
# ('complement', ('identifier', 'a'), ('identifier', 'g'))
# ('complement', ('identifier', 'a'), 'hg19')
target_bed = eval_exp(exp[1], env)
genome = exp[2]
if not type(genome) is str:
genome = eval_exp(genome, env)
return gqltools.complement_bedx(target_bed, genome)
# return gqltools.cast(bedx, new_type)
# }}}
# {{{ elif etype == 'binary-intersect':
elif etype == "binary-intersect":
# print exp
# ('binary-intersect',
# [('identifier', 'a')],
# [('identifier', 'b'), ('identifier', 'c'), ('identifier', 'd')])
idents = exp[1]
n_bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
n_bedfiles = n_bedfiles + [bedx]
n_labels = []
for ident in idents:
n_labels = n_labels + [ident[1]]
idents = exp[2]
m_bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
m_bedfiles = m_bedfiles + [bedx]
m_labels = []
for ident in idents:
m_labels = m_labels + [ident[1]]
return gqltools.binary_intersect_beds(n_bedfiles, n_labels, m_bedfiles, m_labels)
# }}}
# {{{ elif etype == 'binary-jaccard':
elif etype == "binary-jaccard":
# print exp
# ('binary-intersect',
# [('identifier', 'a')],
# [('identifier', 'b'), ('identifier', 'c'), ('identifier', 'd')])
idents = exp[1]
n_bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
n_bedfiles = n_bedfiles + [bedx]
n_labels = []
for ident in idents:
n_labels = n_labels + [ident[1]]
idents = exp[2]
m_bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
m_bedfiles = m_bedfiles + [bedx]
m_labels = []
for ident in idents:
m_labels = m_labels + [ident[1]]
return gqltools.binary_jaccard_beds(n_bedfiles, n_labels, m_bedfiles, m_labels)
# }}}
# {{{ elif etype == 'unary-intersect':
elif etype == "unary-intersect":
# ('intersect',
# [('identifier', 'a'), ('identifier', 'b')])
idents = exp[1]
bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
bedfiles = bedfiles + [bedx]
labels = []
for ident in idents:
labels = labels + [ident[1]]
return gqltools.unary_intersect_beds(bedfiles, labels)
# }}}
# {{{ elif etype == 'subtract':
elif etype == "subtract":
# ('subtract',
# [('identifier', 'a'), ('identifier', 'b')])
ident = exp[1]
bedfile = eval_exp(ident, env)
idents = exp[2]
bedfiles = []
for ident in idents:
bedfiles = bedfiles + [eval_exp(ident, env)]
return gqltools.subtract_beds(bedfile, bedfiles)
# }}}
# {{{ elif etype == 'mergemin':
elif etype == "mergemin":
# print exp
ident_list = exp[1]
if len(ident_list) == 1:
bednfile = eval_exp(ident, env)
result_file = gqltools.mergemin_bedn(bednfile)
return result_file
# }}}
# {{{ elif etype == 'merge':
elif etype == "merge":
# print exp # Debug
# ('merge',
# [('identifier', 'a'), ('identifier', 'b'), ('identifier', 'c')],
# ('score', ('function', 'MIN'), None))
# or
# ('merge',
# [('identifier', 'a'), ('identifier', 'b'), ('identifier', 'c')],
# None)
merge_type = exp[1]
idents = exp[2]
modifiers = exp[3]
bedfiles = []
for ident in idents:
bedx = eval_exp(ident, env)
bedfiles = bedfiles + [bedx]
mods = {}
for modifier in modifiers:
modifier_type = modifier[0]
if modifier_type in mods:
raise InterpException("Multiple definitions of " + modifier_type + ".")
mods[modifier_type] = eval_exp(modifier[1], env)
return gqltools.merge_beds(merge_type, bedfiles, mods)
# }}}
# {{{ elif etype == 'filter':
elif etype == "filter":
# print exp # Debug
# ('filter',
# [ ('identifier', 'a'),
# ('identifier', 'b'),
# ('identifier', 'c'),
# ('identifier', 'd')],
# [ ('score', ('function', 'BOOL',
# [ ('compare', ('<', ('number', 100.0))),
# ('conj', '&'),
# ('compare', ('>', ('number', 50.0))),
# ('conj', '&'),
# ('compare', ('!=', ('number', 75.0)))
# ])
# ),
# ('start', ('function', 'BOOL',
# [ ('compare', ('>', ('number', 1000.0)))
# ])
# ),
# ('end', ('function', 'BOOL',
# [ ('compare', ('<', ('number', 10000.0)))
# ])
# ),
# ('chrom', ('function', 'BOOL',
# [ ('compare', ('==', ('string', 'chr1')))
# ])
# )
# ]
# )
idents = exp[1]
modifiers = exp[2]
bedxs = []
for ident in idents:
bedx = eval_exp(ident, env)
bedxs = bedxs + [bedx]
mods = {}
for modifier in modifiers:
modifier_type = modifier[0]
if modifier_type in mods:
raise InterpException("Multiple definitions of " + modifier_type + ".")
function = eval_exp(modifier[1], env)
if function[0] != "BOOL":
raise InterpException("Unsupported function type in FOREACH: " + function)
bool_funcs = []
for element in function[1]:
bool_func = []
if element[0] == "compare":
op = element[1][0]
val = element[1][1]
if val[0] == "number":
bool_func = [op, val[1]]
elif val[0] == "string":
bool_func = [op, '"' + val[1] + '"']
else:
raise InterpException("Unsupported value type in boolean " + "function in FOREACH.")
elif element[0] == "conj":
conj = element[1][0]
bool_func = [conj]
else:
raise InterpException("Error in boolean function in FOREACH.")
bool_funcs.append(bool_func)
mods[modifier_type] = bool_funcs
return gqltools.filter_bedx(bedxs, mods)
# }}}
# {{{elif etype == 'count':
elif etype == "count":
ident = eval_exp(exp[1], env)
return gqltools.count(ident)
# }}}
# {{{ simple rules
elif etype == "file":
# ('file', ('string', 'file0'))
return eval_exp(exp[1], env)
elif etype == "filetype":
# ('filetype', 'GENOME'))
return gqltypes.source_type_map[exp[1]]
elif etype == "string":
# ('string', 'file0')
return exp[1]
elif etype == "number":
return float(exp[1])
elif etype == "function":
if len(exp) == 2:
return exp[1]
else:
return exp[1:]
else:
print "ERROR: unknown expression type ",
print etype
return None