Merge pull request #134 from dengzq1234/main

correct data_matrix handling

Author: dengzq1234

docs/build/doctrees/documentations.doctree

@@ -23,7 +23,6 @@ TreeProfiler requires
   - scipy >= 1.8.0
   - matplotlib >= 3.4
   - pymc >= 5.0.0
-  - aesara
   - pastml (custom)
 
 Quick install via pip
@@ -34,9 +33,6 @@ Quick install via pip
     pip install --force-reinstall "git+https://github.com/etetoolkit/ete.git@ete4_treeprofiler"
 
 
-    # Install TreeProfiler dependencies
-    pip install biopython selenium scipy matplotlib pymc aesara
-
     # Install custom pastml package for ete4
     pip install "git+https://github.com/dengzq1234/pastml.git@pastml2ete4" 
 

docs/build/doctrees/index.doctree

@@ -232,7 +232,6 @@ <h2><a class="toc-backref" href="#id8" role="doc-backlink">Dependencies</a><a cl
 <li><p>scipy &gt;= 1.8.0</p></li>
 <li><p>matplotlib &gt;= 3.4</p></li>
 <li><p>pymc &gt;= 5.0.0</p></li>
-<li><p>aesara</p></li>
 <li><p>pastml (custom)</p></li>
 </ul>
 </dd>
@@ -244,9 +243,6 @@ <h2><a class="toc-backref" href="#id9" role="doc-backlink">Quick install via pip
 <span class="n">pip</span> <span class="n">install</span> <span class="o">--</span><span class="n">force</span><span class="o">-</span><span class="n">reinstall</span> <span class="s2">&quot;git+https://github.com/etetoolkit/ete.git@ete4_treeprofiler&quot;</span>
 
 
-<span class="c1"># Install TreeProfiler dependencies</span>
-<span class="n">pip</span> <span class="n">install</span> <span class="n">biopython</span> <span class="n">selenium</span> <span class="n">scipy</span> <span class="n">matplotlib</span> <span class="n">pymc</span> <span class="n">aesara</span>
-
 <span class="c1"># Install custom pastml package for ete4</span>
 <span class="n">pip</span> <span class="n">install</span> <span class="s2">&quot;git+https://github.com/dengzq1234/pastml.git@pastml2ete4&quot;</span>
 

docs/build/html/_sources/documentations.rst.txt

@@ -23,7 +23,6 @@ TreeProfiler requires
   - scipy >= 1.8.0
   - matplotlib >= 3.4
   - pymc >= 5.0.0
-  - aesara
   - pastml (custom)
 
 Quick install via pip
@@ -34,9 +33,6 @@ Quick install via pip
     pip install --force-reinstall "git+https://github.com/etetoolkit/ete.git@ete4_treeprofiler"
 
 
-    # Install TreeProfiler dependencies
-    pip install biopython selenium scipy matplotlib pymc aesara
-
     # Install custom pastml package for ete4
     pip install "git+https://github.com/dengzq1234/pastml.git@pastml2ete4" 
 

docs/build/html/documentations.html

@@ -694,7 +694,7 @@ def run_tree_annotate(tree, input_annotated_tree=False,
     return annotated_tree, prop2type
 
 
-def run_array_annotate(tree, array_dict, num_stat='none', column2method={}):
+def run_array_annotate(tree, array_dict, num_stat='none', column2method={}, prop2type={}):
     matrix_props = list(array_dict.keys())
     # annotate to the leaves
     start = time.time()
@@ -708,17 +708,20 @@ def run_array_annotate(tree, array_dict, num_stat='none', column2method={}):
     for node in tree.traverse():
         if not node.is_leaf:
             for prop in matrix_props:
+                
                 # get the array from the children leaf nodes
                 arrays = [child.get_prop(prop) for child in node.leaves() if child.get_prop(prop) is not None]
 
                 if column2method.get(prop) is not None:
                     num_stat = column2method.get(prop)
 
                 stats = compute_matrix_statistics(arrays, num_stat=num_stat)
+                
                 if stats:
                     for stat, value in stats.items():
+                        
                         node.add_prop(utils.add_suffix(prop, stat), value.tolist())
-                        #prop2type[utils.add_suffix(prop, stat)] = float
+                        prop2type[utils.add_suffix(prop, stat)] = list
     end = time.time()
     logger.info(f'Time for run_array_annotate to run: {end - start}')
     return tree
@@ -915,7 +918,7 @@ def run(args):
     )
 
     if args.data_matrix:
-        annotated_tree = run_array_annotate(annotated_tree, array_dict, num_stat=args.num_stat, column2method=column2method)
+        annotated_tree = run_array_annotate(annotated_tree, array_dict, num_stat=args.num_stat, column2method=column2method, prop2type=prop2type)
         # update prop2type
         for filename in array_dict.keys():
             prop2type[filename] = list
@@ -954,12 +957,20 @@ def run(args):
         list_keys = [key for key, value in prop2type.items() if value == list]
         # Replace all commas in the tree with '||'
         list_sep = '||'
-        for node in annotated_tree.leaves():
-            for key in list_keys:
-                if node.props.get(key):
-                    cont2str = list(map(str, node.props.get(key)))
-                    list2str = list_sep.join(cont2str)
-                    node.add_prop(key, list2str)
+        if args.data_matrix and args.num_stat != 'none':
+            for node in annotated_tree.traverse():
+                for key in list_keys:
+                    if node.props.get(key):
+                        cont2str = list(map(str, node.props.get(key)))
+                        list2str = list_sep.join(cont2str)
+                        node.add_prop(key, list2str)
+        else:
+            for node in annotated_tree.leaves():
+                for key in list_keys:
+                    if node.props.get(key):
+                        cont2str = list(map(str, node.props.get(key)))
+                        list2str = list_sep.join(cont2str)
+                        node.add_prop(key, list2str)
 
 
         avail_props = list(prop2type.keys())

docs/build/html/searchindex.js

@@ -2260,11 +2260,13 @@ def process_color_configuration(node2matrix, profiling_props=None):
                 if internal_num_rep != 'none':
                     representative_prop = utils.add_suffix(profiling_prop, internal_num_rep)
                     prop_value = node.props.get(representative_prop)
+                    
                     if prop_value is not None:
                         if isinstance(prop_value, list) or prop2type.get(representative_prop) == list:
                             list_props.add(profiling_prop)
                             if not eteformat_flag:
                                 prop_value = prop_value.split(list_sep)
+                            
                             prop_value = list(map(float, prop_value))
                             if node.name not in node2matrix_list[profiling_prop]:
                                 node2matrix_list[profiling_prop][node.name] = []
@@ -2281,14 +2283,14 @@ def process_color_configuration(node2matrix, profiling_props=None):
     if single_props:
         minval_single, maxval_single, value2color_single, gradientscolor = process_color_configuration(node2matrix_single, profiling_props)
     else:
-        minval_single, maxval_single, value2color_single, gradientscolor = None, None, None
+        minval_single, maxval_single, value2color_single, gradientscolor = None, None, None, None
 
     if list_props:
         # Process list values for each profiling_prop
         results_list = {}
         for prop in profiling_props:
             if prop in list_props:
-                minval_list, maxval_list, value2color_list = process_color_configuration(node2matrix_list[prop], profiling_props)
+                minval_list, maxval_list, value2color_list, gradientscolor = process_color_configuration(node2matrix_list[prop], profiling_props)
                 results_list[prop] = (node2matrix_list[prop], minval_list, maxval_list, value2color_list)
             else:
                 results_list[prop] = (None, None, None, None)