Merge branch 'testing' of https://github.com/DARIAH-DE/Topics into te…

…sting
DARIAH-DE · Feb 22, 2017 · b2054b0 · b2054b0
2 parents 40606a8 + bff3303
commit b2054b0
Show file tree

Hide file tree

Showing 13 changed files with 406 additions and 217,206 deletions.
diff --git a/Introduction.ipynb b/Introduction.ipynb
diff --git a/Mallet.ipynb b/Mallet.ipynb
diff --git a/dariah_topics/mallet.py b/dariah_topics/mallet.py
diff --git a/dariah_topics/visualization.py b/dariah_topics/visualization.py
@@ -262,7 +262,7 @@ def doc_topic_heatmap(data_frame):
     Returns:
 
     """
-    data_frame = data_frame.transpose()
+    data_frame = data_frame.transpose().sort_index()
     doc_labels = list(data_frame.index)
     topic_labels = list(data_frame)
     if len(doc_labels) > 20 or len(topic_labels) > 20: plt.figure(figsize=(20,20))    # if many items, enlarge figure

diff --git a/demonstrator/demonstrator.py b/demonstrator/demonstrator.py
@@ -13,9 +13,11 @@
     https://github.com/DARIAH-DE
 """
 
+import matplotlib
+matplotlib.use('Agg')
 from dariah_topics import preprocessing
-from dariahs_topics import visualization
-from dariahs_topics import mallet
+from dariah_topics import visualization
+from dariah_topics import mallet
 from flask import Flask, request, render_template, send_file
 from gensim.models import LdaModel
 from gensim.corpora import MmCorpus
@@ -68,9 +70,9 @@ def upload_file():
             corpus[label] = tokens
     if 'mallet' in lda:
         print("Creating MALLET binary ...")
-        mallet.create_mallet_model("./mallet_output", "./tmp_files", './mallet/bin/mallet')
+        mallet.create_mallet_model("./mallet_output", "./tmp_files", 'mallet')
         print("Training MALLET LDA model ...")
-        mallet.create_mallet_output('./mallet_output/malletModel.mallet', './mallet_output', './mallet/bin/mallet')
+        mallet.create_mallet_output('./mallet_output/malletModel.mallet', './mallet_output', 'mallet')
         shutil.rmtree('./tmp_files')
         df = mallet.show_topics_keys('./mallet_output')
         doc_topic = mallet.show_docTopicMatrix('./mallet_output')
@@ -127,11 +129,6 @@ def upload_file():
         # Todo: replace by DataFrame.to_html():
         print("Accessing topics for HTML table ...")
         df = visualization.topicwords_in_df(model)
-        import regex
-        pattern = regex.compile(r'\p{L}+\p{P}?\p{L}+')
-        topics = []
-        for n, topic in enumerate(model.show_topics()):
-            topics.append((n+1, pattern.findall(topic[1])))
         print("Rendering result.html ...")
 
         return render_template('result.html', tables=[df.to_html(classes='df')])

diff --git a/demonstrator/templates/result.html b/demonstrator/templates/result.html
@@ -88,7 +88,7 @@
             <div class="span10 offset1 main-content-wrapper no-margin">
                 <div id="content" class="primary-area">
                     <h1>Demonstrator: Topic Modeling</h1>
-                    <div id="contentInner" style="text-align:center; margin:0 auto;">
+                    <div id="contentInner" style="text-align:justify;">
                         {% for table in tables %}
                             {{ table|safe }}
                         {% endfor %}

diff --git a/tutorial_supplementals/mallet_output/doc_topics.txt b/tutorial_supplementals/mallet_output/doc_topics.txt