lab7

.dockerignore

@@ -2,4 +2,7 @@ kaggle.json
 venv
 .vscode
 .idea
-Participants_Data_HPP
+Participants_Data_HPP
+
+my_runs
+saved_model

.gitignore

@@ -220,4 +220,7 @@ venv/*
 
 training_1
 
-Participants_Data_HPP/
+Participants_Data_HPP/
+
+my_runs
+saved_model

Dockerfile

@@ -12,6 +12,7 @@ RUN apt-get install wget
 # RUN python3 -m pip install kaggle
 RUN python3 -m pip install pandas
 RUN pip3 install matplotlib
+RUN pip3 install sacred
 # RUN ln -s ~/.local/bin/kaggle /usr/bin/kaggle
 
 WORKDIR /app

Jenkinsfile3

@@ -15,7 +15,7 @@ pipeline {
                 copyArtifacts projectName: 's444417-create-dataset'
                 sh 'ls -la'
                 sh 'echo $EPOCH_NUMBER'
-                sh 'python3 ./src/trainScript.py $EPOCH_NUMBER'
+                sh 'python3 ./lab7/trainScript.py $EPOCH_NUMBER'
             }
         }
         stage('Archive') {

README.md

@@ -1,6 +1,11 @@
 ## Projekt na przedmiot inżynieria oprogramowania
 IUM_6 opis sposobu rozwiązania zadań i podpunktów
 ---
+Aktualne wyniki zadania IUM_6 dostępne są: 
+- [s444417-create-dataset](https://tzietkiewicz.vm.wmi.amu.edu.pl:8080/job/s444417-create-dataset/): build #244
+- [s444417-training](https://tzietkiewicz.vm.wmi.amu.edu.pl:8080/job/s444417-training/job/master/): build #96
+- [s444417-evaluation](https://tzietkiewicz.vm.wmi.amu.edu.pl:8080/job/s444417-evaluation/job/master/): build #43
+
 Zadanie 1
 1. stworzono job [s444417-training](https://tzietkiewicz.vm.wmi.amu.edu.pl:8080/job/s444417-training/)
 2. s444417-training uruchamia się automatycznie po zakończeniu joba s444417-create-dataset, plik Jenkinsfile, przy pomocy build job. Kopiuje zbiór danych przy pomocy copyArtifact w pliku Jenkinsfile3
@@ -12,8 +17,8 @@ Zadanie 2
 1. stworzono job [s444417-evaluation](https://tzietkiewicz.vm.wmi.amu.edu.pl:8080/job/s444417-evaluation/)
 2. evaluacja modelu metodą evaluate zawołana na modelu w pliku trainScript.py.Zapisanie wyniku do pliku trainResults.csv, w Jenkinsfile.eval archiveArtifact
 3. Jenkinsfile.eval w stagu "Copy prev build artifact" kopiuje trainResults.csv a jeśli go nie ma to catch łapie error, skrypt trainScript.py też obsługuje brak takiego pliku, bo otwiera go w trybie "a+"
-4. skrypt trainScript.py tworzy plota z wczytanych wartości odczytanych z pliku trainResults.csv, natomiast nei ma jak tego podejrzeć w Jenkins
+4. skrypt trainScript.py tworzy plota z wczytanych wartości odczytanych z pliku trainResults.csv i zapisuje wkres do pliku metric.py 
-5.  projekt odpala się po zakończeniu trenowania jenkinsfile3 build job oraz kopiuje sobie model copyArtifacts z uwzględnieniem brancha master
+5. projekt odpala się po zakończeniu trenowania jenkinsfile3 build job oraz kopiuje sobie model copyArtifacts z uwzględnieniem brancha master
 6. copyArtifacts z s444417-create-dataset
 7. parametr BRANCH do wyboru konkretnej gałęzi, buildselector do wybrania builda w Jenkins.eval
-8. powiadomenie mail wysyłane w pliku Jenkinsfile.eval post emailext
+8. powiadomenie mail wraz z metryką loss wysyłane w pliku Jenkinsfile.eval post emailext

lab7/trainScript.py

@@ -0,0 +1,116 @@
+import os
+import sys
+import pandas as pd
+import numpy as np
+
+import tensorflow as tf
+from tensorflow.keras import layers
+
+from sacred import Experiment
+from sacred.observers import FileStorageObserver
+from sacred.observers import MongoObserver
+
+ex = Experiment("sacred_scopes", interactive=True)
+
+ex.observers.append(FileStorageObserver('my_runs'))
+# Mongo observer
+ex.observers.append(MongoObserver(url='mongodb://mongo_user:mongo_password_IUM_2021@localhost:27017', db_name='sacred'))
+
+# train params
+numberOfEpochParam = 0
+
+try: 
+  numberOfEpochParam = int(sys.argv[1])
+except:
+  # dafault val
+  numberOfEpochParam = 3
+
+@ex.config
+def my_config():
+  numberOfEpoch = numberOfEpochParam
+
+@ex.capture
+def train(numberOfEpoch, _run):
+  cwd = os.path.abspath(os.path.dirname(sys.argv[0]))
+
+  pathTrain = cwd + "/../Participants_Data_HPP/Train.csv"
+  pathTest = cwd + "/../Participants_Data_HPP/Test.csv"
+
+  features = ["UNDER_CONSTRUCTION", "RERA", "BHK_NO.", "SQUARE_FT", "READY_TO_MOVE", "RESALE", "LONGITUDE", "LATITUDE", "TARGET(PRICE_IN_LACS)"]
+
+  # get dataset
+  house_price_train = pd.read_csv(pathTrain)[features]
+
+  # get test dataset
+  house_price_test = pd.read_csv(pathTest)[features]
+
+
+  house_price_features = house_price_train.copy()
+  # pop column
+  house_price_labels = house_price_features.pop('TARGET(PRICE_IN_LACS)')
+
+  # process data
+  normalize = layers.Normalization()
+  normalize.adapt(house_price_features)
+
+  feature_test_sample = house_price_test.sample(10)
+  labels_test_sample = feature_test_sample.pop('TARGET(PRICE_IN_LACS)')
+
+  house_price_test_features = house_price_test.copy()
+  # pop column
+  house_price_test_expected = house_price_test_features.pop('TARGET(PRICE_IN_LACS)')
+
+  house_price_features = np.array(house_price_features)
+
+  # load model if exists or create new
+  modelPath = 'saved_model/MyModel_tf'
+  try: 
+    linear_model = tf.keras.models.load_model(modelPath)
+    print("open existing model")
+  except Exception as exception:
+    print(exception)
+    linear_model = tf.keras.Sequential([
+      normalize,
+      layers.Dense(1)
+    ])
+    linear_model.compile(loss = tf.losses.MeanSquaredError(),
+                          optimizer = tf.optimizers.Adam(1))
+    print("creating new model")
+
+  # train model
+  history = linear_model.fit(
+    house_price_features, 
+    house_price_labels, 
+    epochs=int(numberOfEpoch), 
+    validation_split=0.33,
+    verbose=1)
+
+  # save model
+  linear_model.save(modelPath, save_format='tf')
+  # save model as artifact
+  ex.add_artifact(modelPath + "/saved_model.pb")
+
+  # finall loss
+  hist = pd.DataFrame(history.history)
+  hist['epoch'] = history.epoch
+  _run.log_scalar('final.training.loss', hist['loss'].iloc[-1])
+
+  test_results = {}
+  test_results['linear_model'] = linear_model.evaluate(
+      house_price_test_features, house_price_test_expected, verbose=0)
+
+  def flatten(t):
+      return [item for sublist in t for item in sublist]
+
+  pred = np.array(linear_model.predict(feature_test_sample))
+  flatten_pred = flatten(pred)
+
+  with open(cwd + "/../result.txt", "w+") as resultFile:
+    resultFile.write("predictions: " + str(flatten_pred) + '\n')
+    resultFile.write("expected: " + str(labels_test_sample.to_numpy()))
+
+@ex.main
+def main():
+    train()
+
+ex.run()

result.txt

@@ -1,2 +1,2 @@
-predictions: [26.87796, 42.875183, 75.51122, 184.03447, 283.11658, 132.76123, 187.1964, 54.623642, 48.12828, 120.18621]
+predictions: [185.41609, 41.248466, -66.347305, 112.55022, 106.2057, 11.261917, 75.81361, 184.90059, -3.6325989, 85.295105]
-expected: [ 17.   85.   27.  110.  370.   57.9 870.   32.5  76.   38. ]
+expected: [ 96.  51.   8.  63.  25.  11.  80. 110.  85.  41.]

src/trainScript.py

@@ -7,15 +7,6 @@ import tensorflow as tf
 from tensorflow.keras import layers
 import matplotlib.pyplot as plt
 
-def plot_loss(history):
-  plt.plot(history.history['loss'], label='loss')
-  plt.plot(history.history['val_loss'], label='val_loss')
-  plt.xlabel('Epoch')
-  plt.ylabel('Error [MPG]')
-  plt.legend()
-  plt.grid(True)
-  plt.show()
-
 #train params
 numberOfEpoch = sys.argv[1]
 
@@ -85,7 +76,6 @@ history = linear_model.fit(
   verbose=1)
 #callbacks=[cp_callback])
 
-plot_loss(history)
 # save model
 linear_model.save(modelPath, save_format='tf')