Feature/update code for build (#4)

* update pyproject.toml, add readme * update pyproject.toml, call to pydantic * move code around
azuur · Jan 17, 2024 · 890547c · 890547c
1 parent 47916dc
commit 890547c
Show file tree

Hide file tree

Showing 19 changed files with 24 additions and 12 deletions.
diff --git a/.gitignore b/.gitignore
@@ -1,3 +1,6 @@
+# IDE stuff
+.vscode/
+
 # Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]

diff --git a/README.md b/README.md
@@ -0,0 +1,3 @@
+# ML pipelines
+
+Simple ML pipeline repo for experimenting with CI/CD / DevOps / MLOps.
diff --git a/ml_pipelines/common/__init__.py → ml_pipelines/deployment/__init__.py b/ml_pipelines/common/__init__.py → ml_pipelines/deployment/__init__.py
diff --git a/ml_pipelines/eval/__init__.py → ml_pipelines/logic/__init__.py b/ml_pipelines/eval/__init__.py → ml_pipelines/logic/__init__.py
diff --git a/ml_pipelines/serve/__init__.py → ml_pipelines/logic/common/__init__.py b/ml_pipelines/serve/__init__.py → ml_pipelines/logic/common/__init__.py
diff --git a/ml_pipelines/common/dgp.py → ml_pipelines/logic/common/dgp.py b/ml_pipelines/common/dgp.py → ml_pipelines/logic/common/dgp.py
diff --git a/ml_pipelines/common/feature_eng.py → ml_pipelines/logic/common/feature_eng.py b/ml_pipelines/common/feature_eng.py → ml_pipelines/logic/common/feature_eng.py
diff --git a/ml_pipelines/common/model.py → ml_pipelines/logic/common/model.py b/ml_pipelines/common/model.py → ml_pipelines/logic/common/model.py
diff --git a/ml_pipelines/train/__init__.py → ml_pipelines/logic/eval/__init__.py b/ml_pipelines/train/__init__.py → ml_pipelines/logic/eval/__init__.py
diff --git a/ml_pipelines/eval/eval.py → ml_pipelines/logic/eval/eval.py b/ml_pipelines/eval/eval.py → ml_pipelines/logic/eval/eval.py
diff --git a/ml_pipelines/logic/serve/__init__.py b/ml_pipelines/logic/serve/__init__.py
diff --git a/ml_pipelines/serve/serve.py → ml_pipelines/logic/serve/serve.py b/ml_pipelines/serve/serve.py → ml_pipelines/logic/serve/serve.py
diff --git a/ml_pipelines/logic/train/__init__.py b/ml_pipelines/logic/train/__init__.py
diff --git a/ml_pipelines/train/train.py → ml_pipelines/logic/train/train.py b/ml_pipelines/train/train.py → ml_pipelines/logic/train/train.py
diff --git a/ml_pipelines/pipeline/__init__.py b/ml_pipelines/pipeline/__init__.py
diff --git a/ml_pipelines/data_gen_pipeline.py → ml_pipelines/pipeline/data_gen_pipeline.py b/ml_pipelines/data_gen_pipeline.py → ml_pipelines/pipeline/data_gen_pipeline.py
@@ -1,4 +1,4 @@
-from ml_pipelines.common.dgp import generate_raw_data
+from ml_pipelines.logic.common.dgp import generate_raw_data
 
 data = generate_raw_data(10_000, 813)
 data.to_csv("data.csv", index=False)
diff --git a/ml_pipelines/eval_pipeline.py → ml_pipelines/pipeline/eval_pipeline.py b/ml_pipelines/eval_pipeline.py → ml_pipelines/pipeline/eval_pipeline.py
@@ -1,8 +1,8 @@
 import matplotlib.pyplot as plt
 import pandas as pd
 
-from ml_pipelines.common.model import load_model, predict
-from ml_pipelines.eval.eval import (
+from ml_pipelines.logic.common.model import load_model, predict
+from ml_pipelines.logic.eval.eval import (
     calculate_metrics,
     make_roc_plot,
     prob_calibration_plot,

diff --git a/ml_pipelines/train_pipeline.py → ml_pipelines/pipeline/train_pipeline.py b/ml_pipelines/train_pipeline.py → ml_pipelines/pipeline/train_pipeline.py
@@ -1,7 +1,10 @@
 import pandas as pd
 
-from ml_pipelines.common.feature_eng import fit_feature_transform, transform_features
-from ml_pipelines.train.train import save_model, split_data, train_model
+from ml_pipelines.logic.common.feature_eng import (
+    fit_feature_transform,
+    transform_features,
+)
+from ml_pipelines.logic.train.train import save_model, split_data, train_model
 
 # Input
 data = pd.read_csv("data.csv")
@@ -19,4 +22,7 @@
 train_data.to_csv("train_data.csv", index=False)
 test_data.to_csv("test_data.csv", index=False)
 with open("feature_eng_params.json", "w") as f:
-    f.write(feature_eng_params.json())
+    f.write(feature_eng_params.model_dump_json())
+test_data.to_csv("test_data.csv", index=False)
+with open("feature_eng_params.json", "w") as f:
+    f.write(feature_eng_params.model_dump_json())
diff --git a/pyproject.toml b/pyproject.toml
@@ -1,10 +1,14 @@
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"
+
 [tool.poetry]
-name = "mlops-nth-try"
+name = "ml_pipelines"
 version = "0.1.0"
 description = "My nth attempt at making a demo MLOps project"
 authors = ["Adrian Zuur <[email protected]>"]
 readme = "README.md"
-packages = [{include = "mlops_nth_try"}]
+packages = [{include = "ml_pipelines"}]
 
 [tool.poetry.dependencies]
 python = "^3.11"
@@ -22,10 +26,6 @@ pre-commit = "^3.6.0"
 pandas-stubs = "^2.1.4.231227"
 isort = "^5.13.2"
 
-[build-system]
-requires = ["poetry-core"]
-build-backend = "poetry.core.masonry.api"
-
 [tool.ruff]
 line-length = 88
 show-fixes = true
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,3 @@
		# ML pipelines

		Simple ML pipeline repo for experimenting with CI/CD / DevOps / MLOps.