Merge pull request #4 from labdac/last_fm_dataset

last_fm dataset analisis and sessions generation

setup.sh

@@ -1,8 +1,18 @@
 #!/usr/bin/env bash
 
 cd data
+
 wget http://static.echonest.com/millionsongsubset_full.tar.gz
 tar -xvzf millionsongsubset_full.tar.gz
 
 wget https://crossnox.sytes.net/owncloud/index.php/s/edt1ULkBOeQacul/download -O ThirtyMusic.tar.gz
 tar -xvzf ThirtyMusic.tar.gz
+
+#wget http://labrosa.ee.columbia.edu/millionsong/sites/default/files/lastfm/lastfm_train.zip -O lastfm_train.zip
+#unzip -q lastfm_train.zip
+
+#wget http://labrosa.ee.columbia.edu/millionsong/sites/default/files/lastfm/lastfm_test.zip
+#unzip -q lastfm_test.zip
+
+wget http://mtg.upf.edu/static/datasets/last.fm/lastfm-dataset-1K.tar.gz -O lastfm-dataset-1K.tar.gz
+tar -xvzf lastfm-dataset-1K.tar.gz

src/utils/files.py

@@ -0,0 +1,12 @@
+import pickle
+
+
+def save_to_pickle(a, file_name: str):
+    with open(file_name, 'wb') as handle:
+        pickle.dump(a, handle, protocol=pickle.HIGHEST_PROTOCOL)
+
+
+def read_from_pickle(file_name: str):
+    with open(file_name, 'rb') as handle:
+        return pickle.load(handle)
+

src/utils/jupyter.py

@@ -0,0 +1,28 @@
+# Hack to make jupyter notebook wider
+from IPython.display import display, Markdown, HTML
+
+
+def showToggleCodeButton(default_hide=False):
+    code_show = 'true'
+    if not default_hide:
+        code_show = 'false'
+    display(HTML("<style>.container { width:98% !important; }</style>"))
+    display(HTML("<style>.output.output_scroll { height: auto; }</style>"))
+    display(HTML('''<script>
+       code_show=%s;
+       function code_toggle() {
+          if (code_show) {
+              $('div.input').hide();
+              $('div.cell.code_cell:not(:has( .output_wrapper .output_area))').hide();
+          } else {
+              $('div.input').show();
+              $('div.cell.code_cell').show();
+          }
+          code_show = !code_show
+       }
+       $( document ).ready(code_toggle);
+    </script>
+    <form action="javascript:code_toggle()"><input type="submit" class="btn btn-primary" value="Click here to toggle on/off the code."></form>''' % code_show))
+
+def display_mk(mk: str):
+	display(Markdown(mk))

src/utils/pandas.py

@@ -0,0 +1,15 @@
+import pandas as pd
+import copy
+
+class DataFrameWithName(pd.DataFrame):
+    # normal properties
+    _metadata = ['metadata_df_name']
+
+    @property
+    def _constructor(self):
+        return DataFrameWithName
+
+def put_df_name(df, name):
+    df = DataFrameWithName(df)
+    df.metadata_df_name = name
+    return df