obitools/obitools3
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 66 Wiki Activity

obi uniq: added the option to merge ids, except it only works on small

Browse Source 
sets until lists are implemented properly using obiblobs
This commit is contained in:
Celine Mercier
2017-09-25 17:28:03 +02:00
parent 75c15594c4
commit ae24a807da
1 changed files with 36 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
python/obitools3/commands/uniq.pyx
View File

@ -6,7 +6,7 @@ from obitools3.dms.view.view cimport View, Line
from obitools3.dms.view.typed_view.view_NUC_SEQS cimport View_NUC_SEQS
from obitools3.dms.column.column cimport Column, Column_line
from obitools3.dms.capi.obiview cimport QUALITY_COLUMN, COUNT_COLUMN, NUC_SEQUENCE_COLUMN, ID_COLUMN
from obitools3.dms.capi.obitypes cimport OBI_INT, index_t
from obitools3.dms.capi.obitypes cimport OBI_INT, OBI_STR, index_t
from obitools3.apps.optiongroups import addSequenceInputOption, addMinimalOutputOption
from obitools3.uri.decode import open_uri
from obitools3.apps.config import logger
@ -36,7 +36,7 @@ def addOptions(parser):
group.add_argument('--merge-ids', '-e',
action="store_true", dest="uniq:mergeids",
default=False,
help="Add the merged key with all ids of merged sequences.")
help="ONLY WORKING ON SMALL SETS FOR NOW Add the merged key with all ids of merged sequences.")
group.add_argument('--category-attribute', '-c',
action="append", dest="uniq:categories",
@ -55,8 +55,6 @@ def addOptions(parser):
"sequences of any couple of groups are not the"
"prefix of the other one.")
# TODO taxonomy
cdef uniqSequence(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list mergedKeys_list=None, Taxonomy taxonomy=None, bint mergeIds=False, list categories=None) :
@ -81,6 +79,10 @@ cdef uniqSequence(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list
cdef Column_line mcol
cdef Column_line i_mcol
cdef list catl
cdef dict merged_ids_dict
cdef bytes k
cdef int n
#print(categories)
@ -134,6 +136,7 @@ cdef uniqSequence(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list
del(merged_infos)
logger("info", "Second browsing through the input")
merged_ids_dict = {}
i = 0
o_idx = 0
seq_col = view[NUC_SEQUENCE_COLUMN]
@ -199,8 +202,9 @@ cdef uniqSequence(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list
if key_b != COUNT_COLUMN and key_b != ID_COLUMN and key_b != NUC_SEQUENCE_COLUMN and key_b in o_seq and o_seq[key_b] != i_seq[key_b] :
o_seq[key_b] = None
# if mergeIds: # TODO
# u_seq['merged'].append(i_seq.id)
if mergeIds :
merged_ids_dict[o_seq.id].append(i_seq.id)
#o_seq['merged'].append(i_seq.id)
else:
o_view[o_idx] = i_seq
@ -229,10 +233,32 @@ cdef uniqSequence(View_NUC_SEQS view, View_NUC_SEQS o_view, ProgressBar pb, list
mcol[to_merge] = o_seq[COUNT_COLUMN]
o_seq[key] = None # TODO delete column eventually -> make C function?
# if mergeIds:
# u_seq['merged']=[o_seq.id]
if mergeIds:
merged_ids_dict[o_seq.id] = [o_seq.id] # TODO check that this id is added too in the original obiuniq
#o_seq['merged']=[o_seq.id]
i+=1
# Merged ids column
if mergeIds :
nb_ids_max = 0
for k in merged_ids_dict :
n = len(merged_ids_dict[k])
if n > nb_ids_max :
nb_ids_max = n
Column.new_column(o_view,
"merged",
OBI_STR,
nb_elements_per_line=nb_ids_max,
elements_names=None,
comments="obi uniq merged ids",
alias="merged" # TODO what if it already exists
)
for o_seq in o_view:
o_seq['merged'] = merged_ids_dict[o_seq.id]
#TODO
#if taxonomy is not None:
# mergeTaxonomyClassification(uniqSeq, taxonomy)