pip actually does work as specified for mmt_jupyter_kernel, but not installing the binary files yet

interview_kernel/install.py

@@ -3,7 +3,7 @@ import json
 import os
 import sys
 
-from interview_kernel import Interview
+from .interview_kernel import Interview
 
 from jupyter_client.kernelspec import KernelSpecManager
 from IPython.utils.tempdir import TemporaryDirectory

interview_kernel/interview_kernel.py

@@ -13,10 +13,10 @@ from pylatexenc.latex2text import LatexNodes2Text
 import getpass
 from bokeh.io import output_notebook
 
-#from . import pde_state_machine
-import pde_state_machine
-#from . import string_handling
-import string_handling
+from . import pde_state_machine
+#import pde_state_machine
+from . import string_handling
+#import string_handling
 from distutils.util import strtobool
 
 
@@ -320,4 +320,4 @@ Otherwise, you can always answer with \LaTeX-type input.
 if __name__ == '__main__':
     # from ipykernel.kernelapp import IPKernelApp
     # IPKernelApp.launch_instance(kernel_class=Interview)
-    Interview.run_as_main()
\ No newline at end of file
+    Interview.run_as_main()

interview_kernel/pde_state_machine.py

@@ -7,9 +7,9 @@ from collections import OrderedDict
 import getpass
 import re
 
-import string_handling
-from exaoutput import ExaOutput, ExaRunner
-from mmtinterface import *
+from . import string_handling
+from .exaoutput import ExaOutput, ExaRunner
+from .mmtinterface import *
 
 from bokeh.io import output_notebook, show, export_svgs
 from bokeh.plotting import figure