Supporting rarefied tables (#11)

* Supporting rarefied tables

* Fixing travis

* fixing travis (take 2)

* Addressing @ElDeveloper's comment

.travis.yml

@@ -12,25 +12,25 @@ before_install:
   # Update conda itself
   - conda update --yes conda
 install:
-  - travis_retry conda create --yes -n qiita_env python=2.7 pip nose flake8
+  - travis_retry conda create --yes -n qiita python=2.7 pip nose flake8
     pyzmq networkx pyparsing natsort mock future libgfortran
     'pandas>=0.18' 'scipy>0.13.0' 'numpy>=1.7' 'h5py>=2.3.1'
-  - source activate qiita_env
+  - source activate qiita
   - pip install sphinx sphinx-bootstrap-theme coveralls ipython[all]==2.4.1
-  - pip install https://github.com/biocore/qiita/archive/analysis-refactor.zip --process-dependency-links
-  - export MOI_CONFIG_FP=$HOME/miniconda3/envs/qiita_env/lib/python2.7/site-packages/qiita_core/support_files/config_test.cfg
+  - pip install https://github.com/biocore/qiita/archive/master.zip --process-dependency-links
+  - export MOI_CONFIG_FP=$HOME/miniconda3/envs/qiita/lib/python2.7/site-packages/qiita_core/support_files/config_test.cfg
   - ipython profile create qiita-general --parallel
   - qiita-env start_cluster qiita-general
   - qiita-env make --no-load-ontologies
-  - export QIITA_SERVER_CERT=$HOME/miniconda3/envs/qiita_env/lib/python2.7/site-packages/qiita_core/support_files/server.crt
+  - export QIITA_SERVER_CERT=$HOME/miniconda3/envs/qiita/lib/python2.7/site-packages/qiita_core/support_files/server.crt
   - source deactivate
   - travis_retry conda create --yes -n env_name python=$PYTHON_VERSION pip nose flake8 coverage numpy pandas 'h5py>=2.3.1' matplotlib seaborn
   - source activate env_name
   - pip install https://github.com/qiita-spots/qiita_client/archive/master.zip
   - travis_retry pip install . --process-dependency-links
   - configure_biom --env-script "source activate env_name" --server-cert $QIITA_SERVER_CERT
 before_script:
-  - source activate qiita_env
+  - source activate qiita
   - qiita plugins update
   - qiita pet webserver --no-build-docs start &
 script:

qtp_biom/summary.py

@@ -17,39 +17,20 @@
 import seaborn as sns
 
 
-def generate_html_summary(qclient, job_id, parameters, out_dir):
-    """Generates the HTML summary of a BIOM artifact
+def _generate_html(biom):
+    """Generate the HTML summary for the given biom table
 
     Parameters
     ----------
-    qclient : qiita_client.QiitaClient
-        The Qiita server client
-    job_id : str
-        The job id
-    parameters : dict
-        The parameter values to validate and create the artifact
-    out_dir : str
-        The path to the job's output directory
+    biom : biom.Table
+        The biom table to generate the summary from
 
     Returns
     -------
-    bool, None, str
-        Whether the job is successful
-        Ignored
-        The error message, if not successful
+    str
+        A string with the HTML summary page
     """
-    # Step 1: gather file information from qiita using REST api
-    artifact_id = parameters['input_data']
-    qclient_url = "/qiita_db/artifacts/%s/" % artifact_id
-    artifact_info = qclient.get(qclient_url)
-
-    # if we get to this point of the code we are sure that this is a biom file
-    # and that it only has one element
-    fp = artifact_info['files']['biom'][0]
-
-    # Step 2: generate HTML summary
     # Modified from https://goo.gl/cUVHgB
-    biom = load_table(fp)
     num_features, num_samples = biom.shape
 
     sample_counts = []
@@ -64,15 +45,6 @@ def generate_html_summary(qclient, job_id, parameters, out_dir):
         'Median count': np.median(sample_counts),
     }
 
-    ax = sns.distplot(sample_counts)
-    ax.set_xlabel("Number of sequences per sample")
-    ax.set_ylabel("Frequency")
-    plot = ax.get_figure()
-    sc_plot = StringIO()
-    plot.savefig(sc_plot, format='png')
-    sc_plot.seek(0)
-
-    uri = 'data:image/png;base64,' + quote(b64encode(sc_plot.buf))
     artifact_information = [
         "<b>Number of samples:</b> %d<br/>" % num_samples,
         "<b>Number of features:</b> %d<br/>" % num_features,
@@ -84,13 +56,63 @@ def generate_html_summary(qclient, job_id, parameters, out_dir):
          sample_count_summary['Median count']),
         ("<b>Mean count:</b> %d<br/>" %
          sample_count_summary['Mean count']),
-        '<br/><hr/><br/>',
-        '<img src = "%s"/>' % uri
-    ]
+        '<br/><hr/><br/>']
+
+    if sample_count_summary['Minimum count'] == sample_count_summary[
+            'Maximum count']:
+        artifact_information.append(
+            "All the samples in your BIOM table have %d sequences, "
+            "no plot will be shown below."
+            % sample_count_summary['Minimum count'])
+    else:
+        ax = sns.distplot(sample_counts)
+        ax.set_xlabel("Number of sequences per sample")
+        ax.set_ylabel("Frequency")
+        plot = ax.get_figure()
+        sc_plot = StringIO()
+        plot.savefig(sc_plot, format='png')
+        sc_plot.seek(0)
+        uri = 'data:image/png;base64,' + quote(b64encode(sc_plot.buf))
+        artifact_information.append('<img src = "%s"/>' % uri)
+
+    return '\n'.join(artifact_information)
+
+
+def generate_html_summary(qclient, job_id, parameters, out_dir):
+    """Generates the HTML summary of a BIOM artifact
 
+    Parameters
+    ----------
+    qclient : qiita_client.QiitaClient
+        The Qiita server client
+    job_id : str
+        The job id
+    parameters : dict
+        The parameter values to validate and create the artifact
+    out_dir : str
+        The path to the job's output directory
+
+    Returns
+    -------
+    bool, None, str
+        Whether the job is successful
+        Ignored
+        The error message, if not successful
+    """
+    # Step 1: gather file information from qiita using REST api
+    artifact_id = parameters['input_data']
+    qclient_url = "/qiita_db/artifacts/%s/" % artifact_id
+    artifact_info = qclient.get(qclient_url)
+
+    # if we get to this point of the code we are sure that this is a biom file
+    # and that it only has one element
+    fp = artifact_info['files']['biom'][0]
+
+    # Step 2: generate HTML summary
+    biom = load_table(fp)
     of_fp = join(out_dir, "%s.html" % basename(fp))
     with open(of_fp, 'w') as of:
-        of.write('\n'.join(artifact_information))
+        of.write(_generate_html(biom))
 
     # Step 3: add the new file to the artifact using REST api
     success = True

qtp_biom/tests/test_summary.py

@@ -13,9 +13,11 @@
 from shutil import rmtree
 from json import dumps
 
+from biom import Table
+import numpy as np
 from qiita_client.testing import PluginTestCase
 
-from qtp_biom.summary import generate_html_summary
+from qtp_biom.summary import generate_html_summary, _generate_html
 
 
 class SummaryTestsWith(PluginTestCase):
@@ -60,6 +62,13 @@ def test_generate_html_summary(self):
             html = html_f.read()
         self.assertRegexpMatches(html, '\n'.join(EXP_HTML_REGEXP))
 
+    def test_generate_html_summary_rarefied(self):
+        # Create a new biom table
+        data = np.asarray([[0, 2, 4], [2, 2, 2], [4, 2, 0]])
+        table = Table(data, ['O1', 'O2', 'O3'], ['S1', 'S2', 'S3'])
+        obs = _generate_html(table)
+        self.assertEqual(obs, '\n'.join(EXP_HTML_RAREFIED))
+
 
 EXP_HTML_REGEXP = [
     '<b>Number of samples:</b> 7<br/>',
@@ -71,6 +80,17 @@ def test_generate_html_summary(self):
     '<br/><hr/><br/>',
     '<img src = "data:image/png;base64,.*"/>']
 
+EXP_HTML_RAREFIED = [
+    '<b>Number of samples:</b> 3<br/>',
+    '<b>Number of features:</b> 3<br/>',
+    '<b>Minimum count:</b> 6<br/>',
+    '<b>Maximum count:</b> 6<br/>',
+    '<b>Median count:</b> 6<br/>',
+    '<b>Mean count:</b> 6<br/>',
+    '<br/><hr/><br/>',
+    'All the samples in your BIOM table have 6 sequences, '
+    'no plot will be shown below.']
+
 
 if __name__ == '__main__':
     main()