added our best model

Regression_AE5_DSFT.csv

@@ -0,0 +1,2 @@
+Index,Container,Size_Grade,Weight_Kg,Low_Price,High_Price,Sales_Total,Total_Qty_Sold,Total_Kg_Sold,Stock_On_Hand,Day,Month,Year,Province_EASTERN CAPE,Province_NATAL,Province_ORANGE FREE STATE,Province_TRANSVAAL,Province_W.CAPE-BERGRIVER ETC,Province_WEST COAST
+2.0,0,0,18.3,220.0,220.0,1760.0,8.0,146.4,2.0,20.0,1.0,2020.0,0.0,0.0,0.0,0.0,1.0,0.0

api.py

@@ -29,8 +29,7 @@
 
 # Load our model into memory.
 # Please update this path to reflect your own trained model.
-static_model = load_model(
-    path_to_model='assets/trained-models/apples_simple_lm_regression.pkl')
+static_model = load_model(path_to_model='assets/trained-models/best_random.pkl')
 
 print ('-'*40)
 print ('Model succesfully loaded')

model.py

@@ -26,6 +26,7 @@
 import pandas as pd
 import pickle
 import json
+from sklearn import preprocessing
 
 def _preprocess_data(data):
     """Private helper function to preprocess data for model prediction.
@@ -49,6 +50,7 @@ def _preprocess_data(data):
     feature_vector_dict = json.loads(data)
     # Load the dictionary as a Pandas DataFrame.
     feature_vector_df = pd.DataFrame.from_dict([feature_vector_dict])
+
 
     # ---------------------------------------------------------------
     # NOTE: You will need to swap the lines below for your own data
@@ -60,10 +62,19 @@ def _preprocess_data(data):
 
     # ----------- Replace this code with your own preprocessing steps --------
 
-
-    feature_vector_df = feature_vector_df[(feature_vector_df['Commodities'] == 'APPLE GOLDEN DELICIOUS')]
-    predict_vector = feature_vector_df[['Total_Qty_Sold','Stock_On_Hand']]
-
+    #feature_vector_df = feature_vector_df[feature_vector_df['Commodities']  == "APPLE GOLDEN DELICIOUS"]
+    predict_vector = feature_vector_df
+    predict_vector.index.rename('Index',inplace=True)
+    predict_vector.index +=1
+    predict_vector.set_index('Index', inplace=True)
+    label_encoder = preprocessing.LabelEncoder()
+
+    # Encode labels in column 'Size_Grade' in training data. 
+    predict_vector['Size_Grade']= label_encoder.fit_transform(predict_vector['Size_Grade']) 
+
+    # Encode labels in column 'Container' in training data
+    predict_vector['Container']= label_encoder.fit_transform(predict_vector['Container'])   
+    predict_vector = pd.get_dummies(predict_vector, drop_first=True)                   
     # ------------------------------------------------------------------------
 
     return predict_vector
@@ -104,6 +115,7 @@ def make_prediction(data, model):
     """
     # Data preprocessing.
     prep_data = _preprocess_data(data)
+    prep_data.to_csv("Regression_AE5_DSFT.csv")
     # Perform prediction with model and preprocessed data.
     prediction = model.predict(prep_data)
     # Format as list for output standerdisation.

utils/request.py

@@ -21,11 +21,29 @@
 import requests
 import pandas as pd
 import numpy as np
+from sklearn import preprocessing
 
 # Load data from file to send as an API POST request.
 # We prepare a DataFrame with the public test set 
 # from the Kaggle challenge.
-test = pd.read_csv('data/test_data.csv')
+feature_vector_df = pd.read_csv('data/test_data.csv')
+feature_vector_df = feature_vector_df[(feature_vector_df['Commodities'] == 'APPLE GOLDEN DELICIOUS')]
+predict_vector = feature_vector_df
+predict_vector = predict_vector.reset_index(drop=True)
+predict_vector.drop('Commodities',axis = 1, inplace = True)   
+predict_vector['Date'] = predict_vector['Date'].apply(lambda x: pd.to_datetime(x))
+predict_vector['Day'] = predict_vector['Date'].dt.day
+predict_vector['Month'] = predict_vector['Date'].dt.month
+predict_vector['Year'] = predict_vector['Date'].dt.year
+predict_vector.drop('Date', axis = 1, inplace = True)   
+label_encoder = preprocessing.LabelEncoder()
+
+# Encode labels in column 'Size_Grade' in training data. 
+predict_vector['Size_Grade']= label_encoder.fit_transform(predict_vector['Size_Grade']) 
+
+# Encode labels in column 'Container' in training data
+predict_vector['Container']= label_encoder.fit_transform(predict_vector['Container'])   
+test = pd.get_dummies(predict_vector, drop_first=True)   
 
 # Convert our DataFrame to a JSON string.
 # This step is necessary in order to transmit our data via HTTP/S