Fix NDHistogram iteration bug in DpEngine (.items() → .iteritems() / _hist_array) and add Python 3 conda environment file

[sanroman-24]

Description

When running the Cluster module on the python3 branch, the following error occurs:

AttributeError: 'NDHistogram' object has no attribute 'items'

This happens because the code in Cluster/DpEngine.py assumed the input histogram was a Python dict and called .items(). In reality, on the Python 3 branch, the object is a data.Patient.NDHistogram, which exposes .iteritems() (a custom method retained from Python 2 style) and _hist_array.

Fix

Updated DpEngine.init to handle both dict-like inputs and NDHistogram objects.

Added a fallback using .iteritems() to maintain compatibility.

This ensures clustering runs correctly on the Python 3 branch.

Environment

To simplify setup for Python 3 users, I’ve added a conda environment file (phylogicndt_py3.yml) with tested dependencies. This avoids ABI conflicts by using conda for the scientific stack and pip only where required.