U L_#@sXdZddlmZddlmZmZddlmZeeZ Gddde Z Gdd d e Z d S) z, Created on Sep 12, 2012 @author: smirarab )Job)ReadonlySubalignmentExtendedAlignment) get_loggerc@seZdZdZddZddZeeZddZeeZ dd Z ee Z d d Z d d Z ddZddZdddZddZddZddZdS)Problema This class gives a hierarchy of Problems. Each problem includes a list of subproblems, and a pointer to the __parent problem. Root problem has no __parent problem, and tips have empty lists as their __children. This class provides the data structure required to decompose a problem recursively into subproblems. A problem can have many Jobs associated with it. These job objects are kept in a dictionary. cCsFg|_||_|r4t|ts*tdt|||t|_i|_ dS)Nz)Expecting Problem but founding %s instead) _Problem__children_Problem__parent isinstancer TypeErrortype add_childdictjobs annotations)selfparentr+lib/python3.8/site-packages/sepp/problem.py__init__s  zProblem.__init__cCs|jSN)r rrrr get_parent(szProblem.get_parentcCs|jdkrdS|jjdS)Nrr)rlevelrrrr get_level,szProblem.get_levelcCs|jSr)rrrrr get_children0szProblem.get_childrencCs|j|dS)z" Add a child job. N)rappend)rZ subproblemrrrr 4szProblem.add_childccs:t|jdkr|Vn |jD]}|D] }|Vq(qdS)zT Returns an iterator over tips of the problem hierarchy below self. rN)lenr iter_leaves)rctiprrrr:s   zProblem.iter_leavesccs8|j|kr|Vn"|jD]}||D] }|Vq&qdS)z a generator function, returning all Problems under the current problem, with a label tag equal to the given level. Note that this is NOT returning nodes that are 'level' levels below self. N)rriter_nodes_at_level)rrrr rrrr!Es   zProblem.iter_nodes_at_levelcCs||j|<dS)zF A a new job to this problem, to be saved with a name N)r)rjobnamejobrrradd_jobTszProblem.add_jobFc Cs|j|}|dkrdSt|ts0tdt||jrt|dr|jrt|j d}t | }W5QRXt|j d}W5QRX|S|j Snt d|t|fdS)z> returns results of a job, given the job name Nzjob '%s' is not a Job objectresults_on_temprwz#job '%s' has no results set yet: %s) rr rAssertionErrorstrZ result_sethasattrr%openresultevalread Exception)rr"Zignore_missing_resultsr#fresrrrget_job_result_by_nameZs   zProblem.get_job_result_by_namecCsdS)zN returns a label for this Problem. Used in __str__ function. Nrrrrrget_node_labeloszProblem.get_node_labelcCs,|g}|}|jdk r(||j|j}q |Sr)rr)rZretnrrrget_path_to_rootus   zProblem.get_path_to_rootcOs8|}t|jdkr|Sdddd|jD|fS)Nrz(%s)%s,css|]}t|VqdSr)r)).0Zchildrrr sz"Problem.__str__..)r3rrjoin)rargskwargsmerrr__str__}szProblem.__str__N)F)__name__ __module__ __qualname____doc__rrpropertyrrrrZchildrenr rr!r$r2r3r5r=rrrrr s    rc@sxeZdZdZdddZddZeeZddZd d Z eee Z d d Z d dZ ee e Z ddZddZdddZdS) SeppProblemaF A typical Sepp subproblem, defined by a set of taxa, and a parent Adds few basic attributes of a Sepp Problem to the Problem class: 1- a list of taxa associated with this sub problem 2- a Readonly subalignment of the parent problem's alginemnt, induced by this probelms's taxa (with all gaps columns included) 3- the subtree of parent problem's tree, induced by this problem's taxa. This is a deep copy. 4- a set of fragments associated with this subproblem List of taxa is a required field, provided upon initiailzation. Subalignment is automatically computed, and cannot be changed. Subtree is automatically computed, but can be overwriteen by the user Fragments is not automatically computed, and needs to be set by the user NcCs.t||||_d|_d|_d|_d|_dSr)rr_SeppProblem__taxa_SeppProblem__subalignment_SeppProblem__subtree fragmentslabel)rtaxarrrrrs  zSeppProblem.__init__cCs|jSr)rDrrrrget_taxaszSeppProblem.get_taxacCsJ|jdkrDt|jts"t|jdr6t|j|jj|_ntd|j|jS)z If subalignment is not set before, automatically builds a readonly subalignment for this subproblem, based on the taxa assigned to this subproblem. Otherwise, returns the saves subalignment. N subalignmentz.parent object %s has no subalignment attribute) rEr rrCr*rrIrKr rrrrget_subalignments zSeppProblem.get_subalignmentcCs ||_dSr)rE)rZ alignmentrrrset_subalignmentszSeppProblem.set_subalignmentcCsJ|jdkrDt|jts"t|jdr6|jj|j|_ntd|j|jS)z If subtree is not assigned before, automatically buils a subtree based on the taxa assigned to this subproblem. Otherwise, returns the saves subtree. Nsubtreez)parent object %s has no subtree attribute) rFr rrCr*rN get_subtreerIr rrrrrOs zSeppProblem.get_subtreecCs ||_dSr)rF)rrNrrr set_subtreeszSeppProblem.set_subtreecCs|jSr)rHrrrrr3szSeppProblem.get_node_labelcCs@|j}|}||||jd<td|t|f|S)a Writes out the subalignment associated with this subproblem to a file, removing all gap columns, but also keeping track of what was removed and what was kept. This method keeps track of column names that were actually written to file, so that later on, column names could be set to the original value. This is crucial for a correct merge. (see read_extendend_alignment_and_relabel_columns) ref.alignment.columnszHsubalignment without allgap columns written to %s; %d columns remaining.)rKZget_mutable_alignmentZdelete_all_gapZ write_to_pathr_LOGdebugr)rpathZ mut_subalgremaining_colsrrr)write_subalignment_without_allgap_columnss    z5SeppProblem.write_subalignment_without_allgap_columnsTcCsl|jd}|dk rt|dks&tdtd||t|ftt|j}| |||| ||S)a This method goes with write_subalignment_without_allgap_columns method. It enables reading back an alignment that was previously written to disk, and relabeling its columns with the original labels. extension_path is a path to an .sto file (or a list of paths). Alignments from these .sto files are also read, and merged with the original (base) alignment. rQNrzNSubproblem needs to have a proper list of alignment columns associated with itz?Reading %s %s and relabeling it based on %d orig column labels.) rrr(rRrSrlistrGkeysZbuild_extended_alignmentZrelabel_original_columns)r orig_pathZextension_pathZconvert_to_stringrUZap_algrrr,read_extendend_alignment_and_relabel_columnss&    z8SeppProblem.read_extendend_alignment_and_relabel_columns)N)T)r>r?r@rArrJrBrIrLrMrKrOrPrNr3rVrZrrrrrCs   rCN) rAZ schedulerrZsepp.alignmentrrZsepprr>rRobjectrrCrrrrs   x