adding scripts and metadata file

Author: guipleite

dvc.lock

@@ -26,12 +26,12 @@ stages:
       md5: 260904955bdf53e03a72aa2a45fa0297
       size: 4451
     - path: ./src/train.py
-      md5: af34a366856d696c747500bb7b0697ca
-      size: 1093
+      md5: 42f68d29be9da054ca6e6f0e5c142778
+      size: 1094
     outs:
     - path: ./models/model.joblib
-      md5: 033a96ff48de38157b89669b6fdb2a36
-      size: 16901602
+      md5: 287cf6a0a3fabe1103cfaca91bbc9384
+      size: 17122152
   evaluate:
     cmd: python3 ./src/evaluate.py ./data/weatherAUS_processed.csv ./src/model.py
       ./models/model.joblib
@@ -40,21 +40,23 @@ stages:
       md5: 59e89e62fb8f9face4901630d1de3e16
       size: 19507550
     - path: ./models/model.joblib
-      md5: 033a96ff48de38157b89669b6fdb2a36
-      size: 16901602
+      md5: 287cf6a0a3fabe1103cfaca91bbc9384
+      size: 17122152
     - path: ./src/evaluate.py
-      md5: 5fe3fd8936a8571c16b8d5abfe7c4dda
-      size: 878
+      md5: 7e466368d793d09316fc1e078111a9de
+      size: 882
     - path: ./src/model.py
       md5: 260904955bdf53e03a72aa2a45fa0297
       size: 4451
     outs:
     - path: ./results/metrics.json
-      md5: cd5be11e36717182cb48214f5e40e869
-      size: 120
+      md5: 1e190be0481c320f15ff811a21e32490
+      size: 121
     - path: ./results/precision_recall_curve.png
-      md5: 29fb874a238e3e2117429f21d06b6c9d
-      size: 17226
+      md5: d4b1401d5c8601c2f3934e8028785302
+      size: 17033
     - path: ./results/roc_curve.png
-      md5: 8cd591e74fb66be9c8d2b7267ffa3c22
-      size: 20113
+      md5: c38ee7702a4749a5ff60c94a9e575a50
+      size: 20026
+  std_check:
+    cmd: src/scripts/Scripts/std_check.sh ./

dvc.yaml

@@ -1,4 +1,6 @@
 stages:
+  std_check:
+    cmd: src/scripts/Scripts/std_check.sh ./
   preprocess:
     cmd: python3 ./src/preprocess_data.py ./data/weatherAUS.csv
     deps:

metadata.yaml

@@ -0,0 +1,5 @@
+project_name: Rain_aus
+project_version: v0.3
+model_type: scikit-learn_0.23
+author: guipleite
+datetime_creted: 29/03/2021_13:46:23:802394723

results/metrics.json

@@ -1 +1 @@
-{"accuracy": 0.8472656790807144, "recall": 0.965979381443299, "precision": 0.8563279969746628, "f1": 0.9078547325515353}
+{"accuracy": 0.8469057178457704, "recall": 0.9647707074297902, "precision": 0.8567414843577359, "f1": 0.9075526276188406}

src/evaluate.py

@@ -3,6 +3,7 @@
 import pickle
 import os
 import json
+
 # from sklearn.externals import joblib
 import joblib
 
@@ -15,12 +16,14 @@
 
 sys.path.insert(1, MODEL_PATH)
 
+
 def module_from_file(module_name, file_path):
     spec = importlib.util.spec_from_file_location(module_name, file_path)
     module = importlib.util.module_from_spec(spec)
     spec.loader.exec_module(module)
     return module
 
+
 model = module_from_file("model", MODEL_PATH)
 
 # with open(PICKLE_PATH, "rb") as file:
@@ -29,4 +32,4 @@ def module_from_file(module_name, file_path):
 log_eval = model.evaluate(DATA_PATH, pipeline, "./results")
 
 with open("./results/metrics.json", "w") as outfile:
-    json.dump(log_eval["metrics"], outfile)
+    json.dump(log_eval["metrics"], outfile)

src/scripts/Pipelines/README.md

@@ -0,0 +1,16 @@
+python3 model_train_pipeline.py ../../Test_Project/my-model_v1/src/breast_cancer.csv ../../Test_Project/my-model_v1/ 10 ./pickle_model
+
+
+python3 model_deploy_pipeline.py ./pickle_model ../../Test_Project/my-model_v1/ ../../credentials.yaml
+
+
+python3 model_deployed_validate_pipeline.py ../../Test_Project/my-model_v1/src/breast_cancer.csv  ../../credentials.yaml ../../Test_Project/my-model_v1/
+
+
+python3 model_update_pipeline.py ./pickle_model ../../credentials.yaml ../../Test_Project/my-model_v1/
+
+
+python3 model_redeploy_pipeline.py ../../credentials.yaml
+
+
+python3 model_redeploy_pipeline.py ../../credentials.yaml 

src/scripts/Pipelines/model.joblib

@@ -0,0 +1,62 @@
+import os
+import sys
+import pickle
+import yaml
+import joblib
+from ibm_watson_machine_learning import APIClient
+
+MODEL_PATH = os.path.abspath(sys.argv[1])
+PROJ_PATH = os.path.abspath(sys.argv[2])
+CRED_PATH = os.path.abspath(sys.argv[3])
+META_PATH = PROJ_PATH + "/metadata.yaml"
+
+
+with open(CRED_PATH) as stream:
+    try:
+        credentials = yaml.safe_load(stream)
+    except yaml.YAMLError as exc:
+        print(exc)
+
+
+with open(META_PATH) as stream:
+    try:
+        metadata = yaml.safe_load(stream)
+    except yaml.YAMLError as exc:
+        print(exc)
+
+with open(MODEL_PATH, "rb") as file:
+    # pickle_model = pickle.load(file)
+    pipeline = joblib.load(file)
+
+wml_credentials = {"url": credentials["url"], "apikey": credentials["apikey"]}
+
+client = APIClient(wml_credentials)
+client.spaces.list()
+
+MODEL_NAME = metadata["project_name"] + "_" + metadata["project_version"]
+DEPLOY_NAME = MODEL_NAME + "-Deployment"
+MODEL = pipeline
+SPACE_ID = credentials["space_id"]
+
+client.set.default_space(SPACE_ID)
+
+model_props = {
+    client.repository.ModelMetaNames.NAME: MODEL_NAME,
+    client.repository.ModelMetaNames.TYPE: metadata["model_type"],
+    client.repository.ModelMetaNames.SOFTWARE_SPEC_UID: client.software_specifications.get_id_by_name(
+        "default_py3.7"
+    ),
+}
+
+model_details = client.repository.store_model(model=MODEL, meta_props=model_props)
+model_uid = client.repository.get_model_uid(model_details)
+
+deployment_props = {
+    client.deployments.ConfigurationMetaNames.NAME: DEPLOY_NAME,
+    client.deployments.ConfigurationMetaNames.ONLINE: {},
+}
+
+deployment = client.deployments.create(
+    artifact_uid=model_uid,
+    meta_props=deployment_props,
+)

src/scripts/Pipelines/model_deploy_pipeline.py

@@ -0,0 +1,63 @@
+import sys
+import yaml
+import os
+import pandas as pd
+import numpy as np
+from sklearn.metrics import confusion_matrix, accuracy_score
+from sklearn.model_selection import cross_val_score
+from ibm_watson_machine_learning import APIClient
+
+DATA_PATH = os.path.abspath(sys.argv[1])
+CRED_PATH = os.path.abspath(sys.argv[2])
+PROJ_PATH = os.path.abspath(sys.argv[3])
+META_PATH = PROJ_PATH + "/metadata.yaml"
+
+
+def main():
+    with open(CRED_PATH) as stream:
+        try:
+            credentials = yaml.safe_load(stream)
+        except yaml.YAMLError as exc:
+            print(exc)
+
+    data = pd.read_csv(DATA_PATH)
+
+    X = data.iloc[:, :-1]
+    y = data[data.columns[-1]]
+
+    wml_credentials = {"url": credentials["url"], "apikey": credentials["apikey"]}
+
+    client = APIClient(wml_credentials)
+    client.spaces.list()
+
+    SPACE_ID = credentials["space_id"]
+    DEPLOYMENT_UID = input("DEPLOYMENT UID: ")
+
+    client.set.default_space(SPACE_ID)
+
+    # deployment_uid = client.deployments.get_uid(DEPLOYMENT_UID)
+
+    payload = {
+        "input_data": [
+            {"fields": X.columns.to_numpy().tolist(), "values": X.to_numpy().tolist()}
+        ]
+    }
+    result = client.deployments.score(DEPLOYMENT_UID, payload)
+
+    pred_values = np.squeeze(result["predictions"][0]["values"])
+    y_pred_values = [i[0] for i in pred_values]
+
+    def comb_eval(y, y_pred):
+        cm = confusion_matrix(y, y_pred)
+        acc = accuracy_score(y, y_pred)
+
+        return {"cm": cm, "acc": acc}
+
+    eval = comb_eval(y, y_pred_values)
+    print(eval)
+
+    return eval
+
+
+if __name__ == "__main__":
+    main()

src/scripts/Pipelines/model_deployed_validate_pipeline.py

@@ -0,0 +1,35 @@
+import sys
+import importlib.util
+import pickle
+import os
+import json
+
+# from sklearn.externals import joblib
+import joblib
+
+DATA_PATH = os.path.abspath(sys.argv[1])
+# PROJ_PATH = os.path.abspath(sys.argv[2])
+# MODEL_PATH = PROJ_PATH+"/src/model.py"
+MODEL_PATH = sys.argv[2]
+PICKLE_PATH = sys.argv[3]
+
+
+sys.path.insert(1, MODEL_PATH)
+
+
+def module_from_file(module_name, file_path):
+    spec = importlib.util.spec_from_file_location(module_name, file_path)
+    module = importlib.util.module_from_spec(spec)
+    spec.loader.exec_module(module)
+    return module
+
+
+model = module_from_file("model", MODEL_PATH)
+
+# with open(PICKLE_PATH, "rb") as file:
+#         pipeline = pickle.load(file)
+pipeline = joblib.load(PICKLE_PATH)
+log_eval = model.evaluate(DATA_PATH, pipeline, "./results")
+
+with open("./results/metrics.json", "w") as outfile:
+    json.dump(log_eval["metrics"], outfile)