changed requirements.txt for environment.yaml

README.md

@@ -1,5 +1,10 @@
 # treeclust
 
+<p>
+<a href="https://github.com/zeppehpt1/treeclust/blob/main/LICENSE"><img alt="License: MIT" src="https://black.readthedocs.io/en/stable/_static/license.svg"></a>
+<a href="https://github.com/psf/black"><img alt="Code style: black" src="https://img.shields.io/badge/code%20style-black-000000.svg"></a>
+</p>
+
 Python package for automatic processing and clustering of tree species using several different image preprocessing, dimensionality reduction and clustering techniques. The current implementation works with annotated data to take advantage of tree species clustering capabilities based on high-resolution RGB image data. However, with some adjustments, the code can also be used to cluster tree species when tree species labels are not available.
 
 The data used for training the tree detection and delineation model, as well as the data used for the analysis pipeline, can be made available upon request. Please do not hesitate to contact me if you have any questions about the data or individual parts of the process.

docker/dockerfile

@@ -3,8 +3,8 @@ FROM python:3.8
 
 WORKDIR /treeclust
 
-# add req.txt
-COPY requirements.txt .
+# Add dependencies and channel information
+COPY environment.yaml .
 
 # Install base utilities for conda and opencv
 RUN apt-get update \
@@ -23,13 +23,10 @@ RUN wget --quiet https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86
 ENV PATH=$CONDA_DIR/bin:$PATH
 
 RUN conda config --add channels conda-forge \
-    && conda config --add channels pytorch \
-    && conda install mamba pip \
-    && pip install fuzzy-c-means \
-    && ls \
-    && mamba install --file requirements.txt
+    && conda install mamba \
+    && mamba env create -f environment.yaml
 
 # Add files
 COPY . .
 
-CMD ["python", "treeclust/__main__.py"]
+CMD ["conda", "run", "-n", "treeclust", "python", "treeclust/__main__.py"]

environment.yaml

@@ -0,0 +1,36 @@
+name: treeclust
+channels:
+  - conda-forge
+  - pytorch
+  - nvidia
+dependencies:
+  - python=3.8
+  - pip
+  - fiona
+  - geocube
+  - geopandas
+  - joblib
+  - libjpeg-turbo
+  - libpng
+  - matplotlib
+  - numpy
+  - opencv # conda opencv
+  - pandas
+  - pillow
+  - pyarrow
+  - pytest
+  - python-dotenv
+  - rasterio
+  - rioxarray
+  - scikit-learn # conda scikit-learn
+  - scipy
+  - seaborn
+  - shapely
+  - scikit-image
+  - pytorch=2.1.2 # conda pytorch==2.1.2
+  - torchvision=0.16.2
+  - tqdm
+  - umap-learn # conda umap-learn
+  - yellowbrick
+  - pip:
+    - fuzzy-c-means

treeclust/cluster.py

@@ -2,6 +2,8 @@
 import pandas as pd
 import umap as up
 import pickle
+import warnings
+warnings.filterwarnings("ignore", message="n_jobs value -1 overridden to 1 by setting random_state. Use no seed for parallelism.")
 
 from typing import Tuple
 from numpy.typing import NDArray
@@ -32,7 +34,6 @@
 import label_tools as lt
 from constants import NUMBER_OF_CLASSES
 
-
 def load_features(features_path: str) -> Tuple[list, NDArray, list]:
     with open(features_path, "rb") as f:
         data = pickle.load(f)