Don't delete the h5py tests directory.

containers/base/Dockerfile

@@ -72,6 +72,7 @@ RUN echo "deb-src http://ftp.us.debian.org/debian testing main" >> /etc/apt/sour
         futures==3.2.0 \
         google-api-python-client==1.6.2 \
         httplib2==0.10.3 \
+        h5py==2.7.1 \
         ipykernel==4.5.2 \
         ipywidgets==6.0.0 \
         jinja2==2.8 \
@@ -110,8 +111,7 @@ RUN echo "deb-src http://ftp.us.debian.org/debian testing main" >> /etc/apt/sour
         ggplot==0.6.8 \
         google-cloud-dataflow==2.0.0 \
         lime==0.1.1.23 \
-        tensorflow==1.5.0 \
-        h5py==2.7.1 && \
+        tensorflow==1.5.0 && \
     source deactivate && \
 # Clean up before setting up the Python3 env.
     conda clean -tipsy && \
@@ -123,6 +123,7 @@ RUN echo "deb-src http://ftp.us.debian.org/debian testing main" >> /etc/apt/sour
         crcmod==1.7 \
         google-api-python-client==1.6.2 \
         httplib2==0.10.3 \
+        h5py==2.7.1 \
         ipykernel==4.5.2 \
         ipywidgets==6.0.0 \
         jinja2==2.8 \
@@ -161,8 +162,7 @@ RUN echo "deb-src http://ftp.us.debian.org/debian testing main" >> /etc/apt/sour
         bs4==0.0.1 \
         ggplot==0.6.8 \
         lime==0.1.1.23 \
-        tensorflow==1.5.0 \
-        h5py==2.7.1 && \
+        tensorflow==1.5.0 && \
 # Make pip3 a copy of pip for the Python 3 environment.
     cp /usr/local/envs/py3env/bin/pip /usr/local/envs/py3env/bin/pip3 && \
 # Install Python3 IPython kernel
@@ -175,7 +175,9 @@ RUN echo "deb-src http://ftp.us.debian.org/debian testing main" >> /etc/apt/sour
     source deactivate && \
 # Clean up Python packages
     conda clean -tipsy && \
-    find $DATALAB_CONDA_DIR/envs/*/lib -type d -name tests | xargs rm -rf && \
+# Delete tests directories, except for h5py since importing it takes a
+# dependency on it's tests directory.
+    find $DATALAB_CONDA_DIR/envs/*/lib -type d -name tests | grep -v h5py | xargs rm -rf && \
 # Setup Node.js using LTS 6.10
     cd / && \
     mkdir -p /tools/node && \