Sourcery refactored master branch #1
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -10,11 +10,14 @@ | |
|
|
||
| # check if user is just printing version | ||
| if '--version' in argv: | ||
| print(f"TreeCluster version {VERSION}") | ||
| exit() | ||
|
Comment on lines
-13
to
+14
There was a problem hiding this comment. Lines
|
||
|
|
||
| # merge two sorted lists into a sorted list | ||
| def merge_two_sorted_lists(x,y): | ||
| out = [] | ||
| i = 0 | ||
| j = 0 | ||
|
Comment on lines
-17
to
+20
There was a problem hiding this comment. Function
|
||
| while i < len(x) and j < len(y): | ||
| if x[i] < y[j]: | ||
| out.append(x[i]); i+= 1 | ||
|
|
@@ -33,7 +36,7 @@ def merge_multi_sorted_lists(lists): | |
| if len(lists[l]) != 0: | ||
| pq.put((lists[l][0],l)) | ||
| inds = [1 for _ in range(len(lists))] | ||
| out = [] | ||
|
Comment on lines
-36
to
+39
There was a problem hiding this comment. Function
|
||
| while not pq.empty(): | ||
| d,l = pq.get(); out.append(d) | ||
| if inds[l] < len(lists[l]): | ||
|
|
@@ -44,9 +47,9 @@ def merge_multi_sorted_lists(lists): | |
| # get the median of a sorted list | ||
| def median(x): | ||
| if len(x) % 2 != 0: | ||
| return x[len(x) // 2] | ||
| else: | ||
| return (x[len(x) // 2] + x[len(x) // 2 - 1]) / 2 | ||
|
Comment on lines
-47
to
+52
There was a problem hiding this comment. Function
|
||
|
|
||
| # get the average of a list | ||
| def avg(x): | ||
|
|
@@ -59,8 +62,9 @@ def p_to_jc(d,seq_type): | |
|
|
||
| # cut out the current node's subtree (by setting all nodes' DELETED to True) and return list of leaves | ||
| def cut(node): | ||
| cluster = [] | ||
| descendants = Queue() | ||
| descendants.put(node) | ||
|
Comment on lines
-62
to
+67
There was a problem hiding this comment. Function
|
||
| while not descendants.empty(): | ||
| descendant = descendants.get() | ||
| if descendant.DELETED: | ||
|
|
@@ -128,7 +132,7 @@ def pairwise_dists_below_thresh(tree,threshold): | |
| # split leaves into minimum number of clusters such that the maximum leaf pairwise distance is below some threshold | ||
| def min_clusters_threshold_max(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-131
to
+135
There was a problem hiding this comment. Function
|
||
| for node in tree.traverse_postorder(): | ||
| # if I've already been handled, ignore me | ||
| if node.DELETED: | ||
|
|
@@ -179,7 +183,7 @@ def min_clusters_threshold_med_clade(tree,threshold,support): | |
| if node.is_leaf(): | ||
| node.med_pair_dist = 0 | ||
| node.leaf_dists = [0] | ||
| node.pair_dists = [] | ||
|
Comment on lines
-182
to
+186
There was a problem hiding this comment. Function
|
||
| else: | ||
| children = list(node.children) | ||
| l_leaf_dists = [d + children[0].edge_length for d in children[0].leaf_dists] | ||
|
|
@@ -198,7 +202,9 @@ def min_clusters_threshold_med_clade(tree,threshold,support): | |
| del c.leaf_dists; del c.pair_dists | ||
|
|
||
| # perform clustering | ||
| q = Queue() | ||
| q.put(tree.root) | ||
| roots = [] | ||
| while not q.empty(): | ||
| node = q.get() | ||
| if node.med_pair_dist <= threshold: | ||
|
|
@@ -210,7 +216,7 @@ def min_clusters_threshold_med_clade(tree,threshold,support): | |
| # if verbose, print the clades defined by each cluster | ||
| if VERBOSE: | ||
| for root in roots: | ||
| print(f"{root.newick()};", file=stderr) | ||
|
There was a problem hiding this comment. suggestion (llm): Using f-strings is a great improvement for readability and performance. However, ensure that 'stderr' is imported from 'sys' or defined elsewhere in the code. |
||
| return [[str(l) for l in root.traverse_leaves()] for root in roots] | ||
|
|
||
| # average leaf pairwise distance cannot exceed threshold, and clusters must define clades | ||
|
|
@@ -231,7 +237,9 @@ def min_clusters_threshold_avg_clade(tree,threshold,support): | |
| node.avg_pair_dist = node.total_pair_dist/((node.num_leaves*(node.num_leaves-1))/2) | ||
|
|
||
| # perform clustering | ||
| q = Queue() | ||
| q.put(tree.root) | ||
| roots = [] | ||
|
Comment on lines
-234
to
+242
There was a problem hiding this comment. Function
|
||
| while not q.empty(): | ||
| node = q.get() | ||
| if node.avg_pair_dist <= threshold: | ||
|
|
@@ -243,7 +251,7 @@ def min_clusters_threshold_avg_clade(tree,threshold,support): | |
| # if verbose, print the clades defined by each cluster | ||
| if VERBOSE: | ||
| for root in roots: | ||
| print(f"{root.newick()};", file=stderr) | ||
| return [[str(l) for l in root.traverse_leaves()] for root in roots] | ||
|
|
||
| # total branch length cannot exceed threshold, and clusters must define clades | ||
|
|
@@ -258,7 +266,9 @@ def min_clusters_threshold_sum_bl_clade(tree,threshold,support): | |
| node.total_bl = sum(c.total_bl + c.edge_length for c in node.children) | ||
|
|
||
| # perform clustering | ||
| q = Queue() | ||
| q.put(tree.root) | ||
| roots = [] | ||
|
Comment on lines
-261
to
+271
There was a problem hiding this comment. Function
|
||
| while not q.empty(): | ||
| node = q.get() | ||
| if node.total_bl <= threshold: | ||
|
|
@@ -270,13 +280,13 @@ def min_clusters_threshold_sum_bl_clade(tree,threshold,support): | |
| # if verbose, print the clades defined by each cluster | ||
| if VERBOSE: | ||
| for root in roots: | ||
| print(f"{root.newick()};", file=stderr) | ||
| return [[str(l) for l in root.traverse_leaves()] for root in roots] | ||
|
|
||
| # total branch length cannot exceed threshold | ||
| def min_clusters_threshold_sum_bl(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-279
to
+289
There was a problem hiding this comment. Function
|
||
| for node in tree.traverse_postorder(): | ||
| if node.is_leaf(): | ||
| node.left_total = 0; node.right_total = 0 | ||
|
|
@@ -310,7 +320,7 @@ def min_clusters_threshold_sum_bl(tree,threshold,support): | |
| # single-linkage clustering using Metin's cut algorithm | ||
| def single_linkage_cut(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-313
to
+323
There was a problem hiding this comment. Function
|
||
|
|
||
| # find closest leaf below (dist,leaf) | ||
| for node in tree.traverse_postorder(): | ||
|
|
@@ -372,7 +382,7 @@ def single_linkage_cut(tree,threshold,support): | |
| # single-linkage clustering using Niema's union algorithm | ||
| def single_linkage_union(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-375
to
+385
There was a problem hiding this comment. Function
|
||
|
|
||
| # find closest leaf below (dist,leaf) | ||
| for node in tree.traverse_postorder(): | ||
|
|
@@ -432,7 +442,9 @@ def min_clusters_threshold_max_clade(tree,threshold,support): | |
| node.max_pair_dist = max([c.max_pair_dist for c in node.children] + [node.leaf_dist + second_max_leaf_dist]) | ||
|
|
||
| # perform clustering | ||
| q = Queue() | ||
| q.put(tree.root) | ||
| roots = [] | ||
|
Comment on lines
-435
to
+447
There was a problem hiding this comment. Function
|
||
| while not q.empty(): | ||
| node = q.get() | ||
| if node.max_pair_dist <= threshold: | ||
|
|
@@ -444,7 +456,7 @@ def min_clusters_threshold_max_clade(tree,threshold,support): | |
| # if verbose, print the clades defined by each cluster | ||
| if VERBOSE: | ||
| for root in roots: | ||
| print(f"{root.newick()};", file=stderr) | ||
| return [[str(l) for l in root.traverse_leaves()] for root in roots] | ||
|
|
||
| # pick the threshold between 0 and "threshold" that maximizes number of (non-singleton) clusters | ||
|
|
@@ -467,7 +479,7 @@ def argmax_clusters(method,tree,threshold,support): | |
| # cut all branches longer than the threshold | ||
| def length(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-470
to
+482
There was a problem hiding this comment. Function
|
||
| for node in tree.traverse_postorder(): | ||
| # if I've already been handled, ignore me | ||
| if node.DELETED: | ||
|
|
@@ -498,7 +510,9 @@ def length_clade(tree,threshold,support): | |
| node.max_bl = max([c.max_bl for c in node.children] + [c.edge_length for c in node.children]) | ||
|
|
||
| # perform clustering | ||
| q = Queue() | ||
| q.put(tree.root) | ||
| roots = [] | ||
|
Comment on lines
-501
to
+515
There was a problem hiding this comment. Function
|
||
| while not q.empty(): | ||
| node = q.get() | ||
| if node.max_bl <= threshold: | ||
|
|
@@ -510,13 +524,13 @@ def length_clade(tree,threshold,support): | |
| # if verbose, print the clades defined by each cluster | ||
| if VERBOSE: | ||
| for root in roots: | ||
| print(f"{root.newick()};", file=stderr) | ||
| return [[str(l) for l in root.traverse_leaves()] for root in roots] | ||
|
|
||
| # cut tree at threshold distance from root (clusters will be clades by definition) (ignores support threshold if branch is below cutting point) | ||
| def root_dist(tree,threshold,support): | ||
| leaves = prep(tree,support) | ||
| clusters = [] | ||
|
Comment on lines
-519
to
+533
There was a problem hiding this comment. Function
|
||
| for node in tree.traverse_preorder(): | ||
| # if I've already been handled, ignore me | ||
| if node.DELETED: | ||
|
|
@@ -540,7 +554,9 @@ def root_dist(tree,threshold,support): | |
| # cut tree at threshold distance from the leaves (if tree not ultrametric, max = distance from furthest leaf from root, min = distance from closest leaf to root, avg = average of all leaves) | ||
| def leaf_dist(tree,threshold,support,mode): | ||
| modes = {'max':max,'min':min,'avg':avg} | ||
| assert ( | ||
| mode in modes | ||
| ), f"Invalid mode. Must be one of: {', '.join(sorted(modes.keys()))}" | ||
|
Comment on lines
-543
to
+559
There was a problem hiding this comment. Function
|
||
| dist_from_root = modes[mode](d for u,d in tree.distances_from_root(internal=False)) - threshold | ||
| return root_dist(tree,dist_from_root,support) | ||
| def leaf_dist_max(tree,threshold,support): | ||
|
|
@@ -576,13 +592,29 @@ def leaf_dist_avg(tree,threshold,support): | |
| parser.add_argument('-o', '--output', required=False, type=str, default='stdout', help="Output File") | ||
| parser.add_argument('-t', '--threshold', required=True, type=float, help="Length Threshold") | ||
| parser.add_argument('-s', '--support', required=False, type=float, default=float('-inf'), help="Branch Support Threshold") | ||
| parser.add_argument( | ||
| '-m', | ||
| '--method', | ||
| required=False, | ||
| type=str, | ||
| default='max_clade', | ||
| help=f"Clustering Method (options: {', '.join(sorted(METHODS.keys()))})", | ||
| ) | ||
| parser.add_argument( | ||
| '-tf', | ||
| '--threshold_free', | ||
| required=False, | ||
| type=str, | ||
| default=None, | ||
| help=f"Threshold-Free Approach (options: {', '.join(sorted(THRESHOLDFREE.keys()))})", | ||
| ) | ||
| parser.add_argument('-v', '--verbose', action='store_true', help="Verbose Mode") | ||
| parser.add_argument('--version', action='store_true', help="Display Version") | ||
| args = parser.parse_args() | ||
| assert args.method.lower() in METHODS, f"ERROR: Invalid method: {args.method}" | ||
|
There was a problem hiding this comment. thought (llm): The use of f-strings for assertions is good for readability. However, consider the impact on performance if 'METHODS.keys()' is large, as it will be converted to a list and sorted every time this assertion is hit. If performance is a concern, pre-compute the sorted list of method keys outside the assertion. |
||
| assert ( | ||
| args.threshold_free is None or args.threshold_free in THRESHOLDFREE | ||
| ), f"ERROR: Invalid threshold-free approach: {args.threshold_free}" | ||
|
Comment on lines
-579
to
+617
There was a problem hiding this comment. Lines
|
||
| assert args.threshold >= 0, "ERROR: Length threshold must be at least 0" | ||
| assert args.support >= 0 or args.support == float('-inf'), "ERROR: Branch support must be at least 0" | ||
| VERBOSE = args.verbose | ||
|
|
@@ -596,20 +628,23 @@ def leaf_dist_avg(tree,threshold,support): | |
| from sys import stdout; outfile = stdout | ||
| else: | ||
| outfile = open(args.output,'w') | ||
| trees = [] | ||
| for line in infile: | ||
| l = line.decode().strip() if isinstance(line,bytes) else line.strip() | ||
| trees.append(read_tree_newick(l)) | ||
|
|
||
| # run algorithm | ||
| for tree in trees: | ||
| clusters = ( | ||
| METHODS[args.method.lower()](tree, args.threshold, args.support) | ||
| if args.threshold_free is None | ||
| else THRESHOLDFREE[args.threshold_free]( | ||
| METHODS[args.method.lower()], | ||
| tree, | ||
| args.threshold, | ||
| args.support, | ||
| ) | ||
| ) | ||
| outfile.write('SequenceName\tClusterNumber\n') | ||
| cluster_num = 1 | ||
| for cluster in clusters: | ||
|
|
||
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
suggestion (llm): Consider removing the semicolon at the end of the print statement for consistency with the rest of the codebase.