From 32ff78799614aa62412f498d3af27c092d0f356e Mon Sep 17 00:00:00 2001
From: "Borie, Niels" <niels.borie@bleckwen.ai>
Date: Thu, 8 Feb 2024 23:13:07 +0100
Subject: [PATCH] :building_construction: create an advanced image for a given
 input python version

---
 advanced/Dockerfile      | 23 +++++++++++++----------
 advanced/environment.yml | 29 ++++++++++++++++-------------
 2 files changed, 29 insertions(+), 23 deletions(-)

diff --git a/advanced/Dockerfile b/advanced/Dockerfile
index cb3a134..3da9622 100644
--- a/advanced/Dockerfile
+++ b/advanced/Dockerfile
@@ -1,15 +1,18 @@
-# Copyright (c) Niels Borie.
+ARG PYTHON_VERSION
+ARG PYTHON_RELEASE_VERSION
+ARG IMAGE_VERSION
 
-FROM nielsborie/base-py39
+FROM nielsborie/machine-learning-environments:base-py${PYTHON_VERSION}-${IMAGE_VERSION}
 
-LABEL maintainer="Niels BORIE"
+ARG PYTHON_VERSION
+ENV ENV_NAME=py${PYTHON_VERSION}
 
-# --- Install h2o
-RUN pip install -f http://h2o-release.s3.amazonaws.com/h2o/latest_stable_Py.html h2o
-
-# Create the environment:
 COPY environment.yml .
-RUN mamba env update --name py39 --file environment.yml --prune
+RUN mamba env update --name ${ENV_NAME} --file environment.yml --prune
+
+RUN . /opt/conda/etc/profile.d/conda.sh && \
+    conda activate ${ENV_NAME} && \
+    pip install -f http://h2o-release.s3.amazonaws.com/h2o/latest_stable_Py.html h2o && \
+    conda deactivate
 
-# clean up pip cache
-RUN rm -rf /root/.cache/pip/*
\ No newline at end of file
+RUN rm -rf /root/.cache/pip/*
diff --git a/advanced/environment.yml b/advanced/environment.yml
index f73d399..fa95467 100644
--- a/advanced/environment.yml
+++ b/advanced/environment.yml
@@ -1,38 +1,41 @@
-name: py39
 channels:
   - conda-forge
+  - bioconda
+  - defaults
 dependencies:
-  - pip
+  # vanilla
   - xgboost
   - lightgbm
   - catboost
   - mlxtend
-  - gensim
+  - scikit-mdr
   - vowpalwabbit
+  # interpretability
+  - shap
+  - lime
+  # nlp
+  - gensim
+  - spacy
+  # automl
   - hyperopt
+  - optuna
+  - tpot
+  # genetic
   - deap
-  - update_checker
-  - tqdm
-  - stopit
-  - scikit-mdr
   - skrebate
-  - tpot
   - yellowbrick
-  - spacy
   - gplearn
-  - shap
-  - lime
+  # manipulation (dates, GPS, ...)
   - gpxpy
   - arrow
   - haversine
   - toolz
   - cytoolz
   - sacred
-  - plotly
   - missingno
-  - pandas-profiling
   - pip:
       - git+https://github.com/nicta/dora.git
+      - ray[data,train,tune,serve]
       - tables
       - sklearn-deap
       - trueskill