Dev

.dvc/.gitignore

@@ -0,0 +1,3 @@
+/config.local
+/tmp
+/cache

.dvc/config

@@ -0,0 +1,4 @@
+[core]
+    remote = mygd
+['remote "mygd"']
+    url = gdrive://1Kk68L-RBcmvF5xtufnEOc80VIVvsFXof

.dvcignore

@@ -0,0 +1,3 @@
+# Add patterns of files dvc should ignore, which could improve
+# the performance. Learn more at
+# https://dvc.org/doc/user-guide/dvcignore

.github/workflows/main.yml

@@ -1,25 +1,38 @@
-name: model-wine-quality
-on: [push]
+name: CML Report
+on: push
 jobs:
   run:
     runs-on: [ubuntu-latest]
-    container: docker://dvcorg/cml-py3:latest
     steps:
-      - uses: actions/checkout@v2
-      - name: cml_run
+      - uses: iterative/setup-cml@v1
+      - uses: iterative/setup-dvc@v1
+      - uses: actions/checkout@v3
+        with:
+          fetch-depth: 2
+      - name: Generate metrics report
         env:
-          repo_token: ${{ secrets.GITHUB_TOKEN }}
+          REPO_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         run: |
+          if [ $GITHUB_REF = refs/heads/main ]; then
+            PREVIOUS_REF=HEAD~1
+          else
+            PREVIOUS_REF=main
+          fi
 
-          # Your ML workflow goes here
-          pip install -r requirements.txt
-          python train.py
-
-          echo "## Model metrics" > report.md
-          cat metrics.txt >> report.md
-
-          echo "## Data viz" >> report.md
-          cml-publish feature_importance.png --md >> report.md
-          cml-publish residuals.png --md >> report.md
-
-          cml-send-comment report.md
+          echo "# CML Report" > report.md
+          echo "## Plots" >> report.md
+          dvc plots diff $PREVIOUS_REF workspace \
+            --show-vega --targets evaluation/plots/precision_recall.json > vega.json
+          vl2svg vega.json prc.svg
+          cml publish prc.svg --title "Precision & Recall" --md >> report.md
+
+          dvc plots diff $PREVIOUS_REF workspace \
+            --show-vega --targets evaluation/plots/confusion_matrix.json > vega.json
+          vl2svg vega.json confusion.svg
+          cml publish confusion.svg --title "Confusion Matrix" --md >> report.md
+
+          echo "## Metrics and Params" >> report.md
+          echo "### $PREVIOUS_REF → workspace" >> report.md
+          dvc exp diff $PREVIOUS_REF --show-md >> report.md
+
+          cml send-comment report.md

.gitignore

@@ -0,0 +1 @@
+/model.pkl

data/.gitignore

@@ -0,0 +1,3 @@
+/data.xml
+/prepared
+/features

data/data.xml.dvc

@@ -0,0 +1,12 @@
+md5: 8ad921bd158a9b447da32d275f72eb98
+frozen: true
+deps:
+- path: data.xml
+  repo:
+    url: https://github.com/NerdToMars/wine_data.git
+    rev: v0.1
+    rev_lock: e6ed3c40ac4f03d9387791da1ced5acced35ae9b
+outs:
+- md5: 079fbd15fa2c32c539c4c4e3675b514a
+  size: 28890194
+  path: data.xml

dvc.lock

@@ -0,0 +1,93 @@
+schema: '2.0'
+stages:
+  prepare:
+    cmd: python src/prepare.py data/data.xml
+    deps:
+    - path: data/data.xml
+      md5: 079fbd15fa2c32c539c4c4e3675b514a
+      size: 28890194
+    - path: src/prepare.py
+      md5: f09ea0c15980b43010257ccb9f0055e2
+      size: 1576
+    params:
+      params.yaml:
+        prepare.seed: 20170428
+        prepare.split: 0.2
+    outs:
+    - path: data/prepared
+      md5: 2fe72f304d64c28867d884e798568460.dir
+      size: 16874726
+      nfiles: 2
+  featurize:
+    cmd: python src/featurization.py data/prepared data/features
+    deps:
+    - path: data/prepared
+      md5: 2fe72f304d64c28867d884e798568460.dir
+      size: 16874726
+      nfiles: 2
+    - path: src/featurization.py
+      md5: e0265fc22f056a4b86d85c3056bc2894
+      size: 2490
+    params:
+      params.yaml:
+        featurize.max_features: 100
+        featurize.ngrams: 2
+    outs:
+    - path: data/features
+      md5: 9202bf271b063e36efe33700e553fcef.dir
+      size: 3295678
+      nfiles: 2
+  train:
+    cmd: python src/train.py data/features model.pkl
+    deps:
+    - path: data/features
+      md5: 9202bf271b063e36efe33700e553fcef.dir
+      size: 3295678
+      nfiles: 2
+    - path: src/train.py
+      md5: c3961d777cfbd7727f9fde4851896006
+      size: 967
+    params:
+      params.yaml:
+        train.min_split: 0.01
+        train.n_est: 50
+        train.seed: 20170428
+    outs:
+    - path: model.pkl
+      md5: 71fa4c5d016f2fc1c00a6fd3d2385e57
+      size: 1874902
+      isexec: true
+  evaluate:
+    cmd: python src/evaluate.py model.pkl data/features
+    deps:
+    - path: data/features
+      md5: 9202bf271b063e36efe33700e553fcef.dir
+      size: 3295678
+      nfiles: 2
+    - path: model.pkl
+      md5: 71fa4c5d016f2fc1c00a6fd3d2385e57
+      size: 1874902
+    - path: src/evaluate.py
+      md5: 44e714021a65edf881b1716e791d7f59
+      size: 2346
+    outs:
+    - path: evaluation.json
+      md5: b07f807e73ad5e67347a6956c671c3aa
+      size: 72
+      isexec: true
+    - path: evaluation/importance.png
+      md5: 629d6f80e9fc5cfa2b3d872a4a5cbd1c
+      size: 32683
+      isexec: true
+    - path: evaluation/plots/confusion_matrix.json
+      md5: b99494fb5d522a16def52938bef636c2
+      size: 243331
+      isexec: true
+    - path: evaluation/plots/precision_recall.json
+      md5: 1a8e2a78fa5608d46c985ea850276598
+      size: 133445
+      isexec: true
+    - path: evaluation/plots/roc.json
+      md5: e88b258d360279efa3eab6b68716a874
+      size: 159030
+      isexec: true

dvc.yaml

@@ -0,0 +1,56 @@
+stages:
+  prepare:
+    cmd: python src/prepare.py data/data.xml
+    deps:
+    - data/data.xml
+    - src/prepare.py
+    params:
+    - prepare.seed
+    - prepare.split
+    outs:
+    - data/prepared
+  featurize:
+    cmd: python src/featurization.py data/prepared data/features
+    deps:
+    - data/prepared
+    - src/featurization.py
+    params:
+    - featurize.max_features
+    - featurize.ngrams
+    outs:
+    - data/features
+  train:
+    cmd: python src/train.py data/features model.pkl
+    deps:
+    - data/features
+    - src/train.py
+    params:
+    - train.min_split
+    - train.n_est
+    - train.seed
+    outs:
+    - model.pkl
+  evaluate:
+    cmd: python src/evaluate.py model.pkl data/features
+    deps:
+    - data/features
+    - model.pkl
+    - src/evaluate.py
+    metrics:
+    - evaluation.json:
+        cache: false
+    plots:
+    - evaluation/importance.png
+    - evaluation/plots/confusion_matrix.json:
+        cache: false
+        template: confusion
+        x: actual
+        y: predicted
+    - evaluation/plots/precision_recall.json:
+        cache: false
+        x: recall
+        y: precision
+    - evaluation/plots/roc.json:
+        cache: false
+        x: fpr
+        y: tpr

evaluation.json

@@ -0,0 +1,4 @@
+{
+    "avg_prec": 0.916698900628581,
+    "roc_auc": 0.9508481917506465
+}

evaluation/.gitignore

@@ -0,0 +1 @@
+/importance.png