diff --git a/api/__pycache__/app.cpython-310.pyc b/api/__pycache__/app.cpython-310.pyc
index e0cb02c71747bd6c7c0edfdb70161572f9f8e08a..a694bd589ce1602823bdfe06ffd3bb86cfc44283 100644
Binary files a/api/__pycache__/app.cpython-310.pyc and b/api/__pycache__/app.cpython-310.pyc differ
diff --git a/api/run.py b/api/run.py
index 8e1ba814547677540c27df9ee15c8bf9469236d8..71116980b90ad7cae6b5de6588a9e270e8f543f2 100644
--- a/api/run.py
+++ b/api/run.py
@@ -10,10 +10,11 @@ import pandas as pd
import datetime as dt
import numpy as np
import json
+import time
from sqlalchemy import create_engine
-from sklearn.preprocessing import MinMaxScaler
+from sklearn.preprocessing import MinMaxScaler
from tensorflow import keras
from keras.models import Sequential
from keras.layers import Dense, LSTM, Dropout
@@ -48,12 +49,23 @@ x_test = np.array([])
y_test = np.array([])
predictions = np.array([])
-# Lstm Model
+
+# Initialising the LSTM model with training
+
model = Sequential()
-feature_length = 100
+feature_length = 10
+testing_records=5
+batch_size = 5
+epochs = 10
+accuracy=0
+execute_time=0
# Scaler
-scaler = MinMaxScaler(feature_range=(0,1))
+scaler = MinMaxScaler()
+
+# MySQL database connection
+db_connection_str = 'mysql+pymysql://root:@localhost/csct'
+db_connection = create_engine(db_connection_str)
#============================================
@@ -63,15 +75,15 @@ scaler = MinMaxScaler(feature_range=(0,1))
# Read data from data base and re arrange data
+#---------------------------------------------
def read_data_set():
# Declare global variable
global data_set
global raw_data_set
global index_data_set
- # Define database connection
- db_connection_str = 'mysql+pymysql://root:@localhost/csct'
- db_connection = create_engine(db_connection_str)
+
+
# Read data in to Data Frame
raw_data_set = pd.read_sql('SELECT * FROM aw_product_demand', con=db_connection)
@@ -95,35 +107,61 @@ def read_data_set():
return data_set
+#-------------------------------------------------------------------------------
+def save_tranning_history(epochs,time_steps,batch_size,execute_time,accuracy):
+ id=db_connection.execute("INSERT INTO `trannings` ( `date_time`, `epochs`, `time_steps`, `batch_size`, `execute_time`, `accurecy`) VALUES ( CURRENT_TIMESTAMP, '"+str(epochs)+"', '"+str(time_steps)+"', '"+str(batch_size)+"', '"+str(execute_time)+"', '"+str(accuracy)+"')")
+ print("Row Added = ",id.rowcount)
+
+#-------------------------------------------------------------------------------
+
# Function to create x and y data
-def Create_Features_and_Targets(data, feature_length):
- # Declare Xand Y List
- X = list()
- Y = list()
- for i in range(len(data) - feature_length -1):
- # create X array shift by feature length
- X.append(data[i:(i + feature_length), 0])
-
- Y.append(data[i + feature_length, 0])
- # convert to numpy array format
- X = np.array(X)
- Y = np.array(Y)
- return X,Y
+def input_and_targert(data,feature_length):
+
+ # Genarate Samples
+ x_samples = list()
+ y_samples = list()
+ NumerOfRows=len(data)
+ for i in range(feature_length , NumerOfRows , 1):
+ x_sample = data[i-feature_length:i]
+ y_sample = data[i]
+ x_samples.append(x_sample)
+ y_samples.append(y_sample)
+
+ # Reshape the input as a 3D (Number of samles,length of features,features)
+
+ #Reshape input
+ X = np.array(x_samples)
+ X=X.reshape(X.shape[0],X.shape[1],1)
+ print("\n____Input Data Shape :____")
+ print(X.shape)
+
+ # Reshape Target
+ Y=np.array(y_samples)
+ Y=Y.reshape(Y.shape[0],1)
+ print("\n ______Target Data Shape :______")
+ print(Y.shape)
+
+ return X,Y
+#----------------------------------------------------------------------------
# function to calculate prediction
def predict_given_date(data, date, feature_length):
if date not in data.index:
data.loc[date]=0
idx = data.index.get_loc(date)
- close_col = data.iloc[:,1:2]
- close_col = close_col.iloc[idx - feature_length : idx,:].values
- close_col = np.expand_dims(scaler.transform(close_col) , axis = 0)
- Prediction = model.predict(close_col)
- Prediction=np.array(Prediction).reshape(-1, 1)
- Prediction = scaler.inverse_transform(Prediction)
- return Prediction
-# =========================== End of Function =================
+ # close_col = data.iloc[:,1:2]
+ # close_col = close_col.iloc[idx - feature_length : idx,:].values
+ # close_col = np.expand_dims(scaler.transform(close_col) , axis = 0)
+ # Prediction = model.predict(close_col)
+ # Prediction=np.array(Prediction).reshape(-1, 1)
+ # Prediction = scaler.inverse_transform(Prediction)
+ return idx
+
+
+#------------------------------------------------------------------------------
+
+# =========================== End of support Functions =================
#============================= Main funtion ===================
def setup():
@@ -133,6 +171,9 @@ def setup():
global data_set
global predictions
global result
+ global scaler
+ global accuracy
+ global execute_time
# Read Data from DataSet
read_data_set()
@@ -142,57 +183,114 @@ def setup():
data = data_set.iloc[:,1:2]
data = data.values
+ ds = scaler.fit(data)
+ scal_data=ds.transform(data)
+
+
+ scal_data = scaler.fit_transform(data)
+
+ x,y= input_and_targert(scal_data,feature_length)
+
+ # Split the data into training and test
+ x_train=x[:-testing_records]
+ x_test=x[-testing_records:]
+ y_train=y[:-testing_records]
+ y_test=y[-testing_records:]
+
+
+ print("\n___Tranning Data Shape___ ")
+ print(x_train.shape)
+ print(y_train.shape)
+ print("\n___Testing Data Shape___ ")
+ print(x_test.shape)
+ print(y_test.shape)
+
+ for inp, out in zip(x_train[0:2], y_train[0:2]):
+ print(inp,'--', out)
- data = scaler.fit_transform(data)
-
-
- X_train,y_train= Create_Features_and_Targets(data,feature_length)
- X_train = np.reshape(X_train,(X_train.shape[0],X_train.shape[1],1))
-
- print(X_train)
-
- # model
- model = Sequential([
- LSTM(100,return_sequences=True,input_shape=(X_train.shape[1],1)),
- Dropout(0.3),
- LSTM(100, return_sequences = False),
- Dropout(0.3),
-
- Dense(1),
- ])
- model.compile(optimizer='adam',loss="mean_squared_error")
- model.summary()
-
- # Training the model
- history = model.fit(
- X_train,
- y_train,
- epochs = 5,
- batch_size = 12,
- verbose=1,
- )
-
- testData = data_set.iloc[:,1:2] # Get 'Close' feature
- y_real=testData.iloc[feature_length+1:,0:].values #Actual values
- x_test = testData.iloc[:,0:].values # data to test
- # normalizing the Data using Scaler.transform function
- x_test = scaler.transform(x_test)
- x_test, y_test = Create_Features_and_Targets(x_test, feature_length)
- # Making data 3 dimensional
- x_test = np.reshape(x_test,(x_test.shape[0],x_test.shape[1],1))
- y_pred = model.predict(x_test)
- predictions=np.array(scaler.inverse_transform(y_pred)).ravel()
- result["Date"]=data_set.iloc[feature_length+1:]["Date"]
- result["Predictions"]=predictions
- result["OrderDemand"]=data_set.iloc[feature_length+1:]["Order_Demand"]
-
- result.plot()
- plt.show()
- print(result)
-
- #save model
- model.save('lstm_model.h5')
+ # Defining Input shapes for LSTM
+ time_steps=x_train.shape[1]
+ features=x_train.shape[2]
+ print("Number of TimeSteps:", time_steps)
+ print("Number of Features:", features)
+
+
+
+ #Add First LSTM Layer
+
+ model.add(LSTM(units = 10, activation = 'relu', input_shape = (time_steps, features), return_sequences=True))
+
+ # Adding the Second hidden layer and the LSTM layer
+
+ model.add(LSTM(units = 5, activation = 'relu', input_shape = (time_steps, features), return_sequences=True))
+
+ # Adding the Third hidden layer and the LSTM layer
+ model.add(LSTM(units = 5, activation = 'relu', return_sequences=False ))
+
+
+ # Adding the output layer
+ model.add(Dense(units = 1))
+
+ # Compiling model
+ model.compile(optimizer = 'adam', loss = 'mean_squared_error')
+
+
+ # Measuring the time taken by the model to train
+ start_time=time.time()
+
+ # Fitting the RNN to the Training set
+ model_history=model.fit(x_train, y_train, batch_size , epochs)
+
+ end_time=time.time()
+ execute_time=round((end_time-start_time)/60,2)
+ print("__Total Time Taken: ", execute_time, ' Minutes___')
+
+
+
+ # Making predictions on test data
+ predicted_Price = model.predict(x_test)
+ predicted_Price = scaler.inverse_transform(predicted_Price)
+
+ # Getting the original price values for testing data
+ orig=y_test
+ orig=scaler.inverse_transform(y_test)
+
+ # Accuracy of the predictions
+ accuracy=100 - (100*(abs(orig-predicted_Price)/orig)).mean()
+ accuracy=round(accuracy,2)
+ print('Accuracy:',accuracy)
+
+ save_tranning_history(epochs,time_steps,batch_size,execute_time,accuracy)
+
+
+ # Generating predictions on full data
+ TrainPredictions=scaler.inverse_transform(model.predict(x_train))
+ TestPredictions=scaler.inverse_transform(model.predict(x_test))
+
+ FullDataPredictions=np.append(TrainPredictions, TestPredictions)
+ FullDataOrig=data[time_steps:]
+
+ # plotting the full data
+ plt.plot(FullDataPredictions, color = 'blue', label = 'Predicted Price')
+ plt.plot(FullDataOrig , color = 'lightblue', label = 'Original Price')
+
+ plt.title('Stock Price Predictions')
+ plt.xlabel('Trading Date')
+ plt.ylabel('Stock Price')
+ plt.legend()
+ fig=plt.gcf()
+ fig.set_figwidth(20)
+ fig.set_figheight(8)
+ plt.show()
+
+
+ result["Date"]=data_set.iloc[time_steps:]["Date"]
+ result["Predictions"]=FullDataPredictions
+ result["OrderDemand"]=data[time_steps:]
+
+ print(data_set.tail(10))
+
# ====================================================================
# main function call
@@ -293,20 +391,20 @@ def forecast_to_date():
result=predict_given_date(index_data_set,new_date, feature_length)
- df=pd.DataFrame()
- # df=pd.DataFrame(data=result,columns=["Prediction"])
- df['Date']=pd.date_range(start=new_date,periods=feature_length)
- df=df.loc[::-1]
- df['Prediction']=result
+ # df=pd.DataFrame()
+ # # df=pd.DataFrame(data=result,columns=["Prediction"])
+ # df['Date']=pd.date_range(start=new_date,periods=feature_length)
+ # df=df.loc[::-1]
+ # df['Prediction']=result
- df['Date']= pd.to_datetime(df['Date']).dt.date
- df.sort_values('Date', inplace=True)
- df['Date']=df['Date'].astype(str)
- df=df.set_index(df['Date'])
- df=df.tail(1)
+ # df['Date']= pd.to_datetime(df['Date']).dt.date
+ # df.sort_values('Date', inplace=True)
+ # df['Date']=df['Date'].astype(str)
+ # df=df.set_index(df['Date'])
+ # df=df.tail(1)
- return df.to_json(orient='records')
+ return result
@app.route("/forecast_to_range",methods=["POST"])
def forecast_to_range():
diff --git a/dash_board.php b/dash_board.php
index 8ea5be195d9a6d35a75078233effdfd5ff223d18..e61ea547981413d010af6a99c32220e103591da8 100644
--- a/dash_board.php
+++ b/dash_board.php
@@ -14,6 +14,8 @@
<?php include_once './data/result_data_set.php';?>
+ <?php if($date_validate!=''){?>
+
<div class="row">
<div class="col-md-12">
@@ -34,6 +36,8 @@
</div>
+ <?php }?>
+
<script type="text/javascript">
@@ -41,180 +45,59 @@
var myChart = echarts.init(chartDom);
var option;
-
- const upColor = '#ec0000';
- const upBorderColor = '#8A0000';
- const downColor = '#00da3c';
- const downBorderColor = '#008F28';
- // Each item: open，close，lowest，highest
- const data0 = splitData([
- <?=$train_set?>
- ]);
-
- function splitData(rawData) {
- const categoryData = [];
- const values = [];
- for (var i = 0; i < rawData.length; i++) {
- categoryData.push(rawData[i].splice(0, 1)[0]);
- values.push(rawData[i]);
- }
- return {
- categoryData: categoryData,
- values: values
- };
- }
-
- function calculateMA(dayCount) {
- var result = [];
- for (var i = 0, len = data0.values.length; i < len; i++) {
- if (i < dayCount) {
- result.push('-');
- continue;
- }
- var sum = 0;
- for (var j = 0; j < dayCount; j++) {
- sum += +data0.values[i - j][1];
- }
- result.push(sum / dayCount);
- }
- return result;
- }
option = {
title: {
- text: 'Product Demand',
- left: 0
+ text: 'Order Demad'
},
tooltip: {
- trigger: 'axis',
- axisPointer: {
- type: 'cross'
- }
+ trigger: 'axis'
},
legend: {
- data: ['Product Demand']
+ data: ['Order Demand', 'Predictions', 'Error']
},
grid: {
- left: '10%',
- right: '10%',
- bottom: '15%'
+ left: '3%',
+ right: '4%',
+ bottom: '3%',
+ containLabel: true
+ },
+ toolbox: {
+ feature: {
+ saveAsImage: {}
+ }
},
xAxis: {
type: 'category',
- data: data0.categoryData,
- boundaryGap: false,
- axisLine: {
- onZero: false
- },
- splitLine: {
- show: false
- },
- min: 'dataMin',
- max: 'dataMax'
+
+ data: [<?=$date_validate?>]
},
yAxis: {
- scale: true,
- splitArea: {
- show: true
- }
+ type: 'value'
},
- dataZoom: [{
- type: 'inside',
- start: 50,
- end: 100
+ series: [{
+ name: 'Order Deamand',
+ type: 'line',
+
+ data: [<?=$order_validate?>]
},
{
- show: true,
- type: 'slider',
- top: '90%',
- start: 50,
- end: 100
- }
- ],
- series: [{
- name: 'Product Demand',
- type: 'candlestick',
- data: data0.values,
- itemStyle: {
- color: upColor,
- color0: downColor,
- borderColor: upBorderColor,
- borderColor0: downBorderColor
- },
- markPoint: {
- label: {
- formatter: function(param) {
- return param != null ? Math.round(param.value) + '' : '';
- }
- },
- data: [
-
- {
- name: 'average value on close',
- type: 'average',
- valueDim: 'close'
- }
- ],
- tooltip: {
- formatter: function(param) {
- return param.name + '<br>' + (param.data.coord || '');
- }
- }
- },
- markLine: {
- symbol: ['none', 'none'],
- data: [
- [{
- name: 'from lowest to highest',
- type: 'min',
- valueDim: 'lowest',
- symbol: 'circle',
- symbolSize: 10,
- label: {
- show: false
- },
- emphasis: {
- label: {
- show: false
- }
- }
- },
- {
- type: 'max',
- valueDim: 'highest',
- symbol: 'circle',
- symbolSize: 10,
- label: {
- show: false
- },
- emphasis: {
- label: {
- show: false
- }
- }
- }
- ],
- {
- name: 'min line on close',
- type: 'min',
- valueDim: 'close'
- }
- ]
- }
+ name: 'Predections',
+ type: 'line',
+
+ data: [<?=$predection?>]
},
{
- name: 'MA5',
+ name: 'Error',
type: 'line',
- data: calculateMA(5),
- smooth: true,
- lineStyle: {
- opacity: 0.5
- }
+
+ data: [<?=$error?>]
}
]
};
-
option && myChart.setOption(option);
</script>
+
+
<script src="dist/js/chart2.js" type="text/javascript"> </script>
<script src="dist/js/chart3.js" type="text/javascript"> </script>
diff --git a/data/result_data_set.php b/data/result_data_set.php
index b9f1daeb455db89c0a72ef479850d563d267334b..2774487034aad7190774f7b5c3b262b13a83b2c5 100644
--- a/data/result_data_set.php
+++ b/data/result_data_set.php
@@ -1,9 +1,11 @@
<?php
+// Fetch Result from API Core
+
$curl = curl_init();
curl_setopt_array($curl, array(
- CURLOPT_URL => 'http://127.0.0.1:5000/result_tranning',
+ CURLOPT_URL => 'http://127.0.0.1:5000/validation',
CURLOPT_RETURNTRANSFER => true,
CURLOPT_ENCODING => '',
CURLOPT_MAXREDIRS => 10,
@@ -13,40 +15,7 @@ curl_setopt_array($curl, array(
CURLOPT_CUSTOMREQUEST => 'GET',
));
-$training_data = curl_exec($curl);
-
-curl_close($curl);
-
-$arr = json_decode($training_data, true);
-
-$dates = '';
-$order_demands = '';
-$train_set = '';
-
-
-foreach ($arr as $key => $value) {
- $dates = $dates."'".$value['Date']."',";
- $order_demands= $order_demands."'".$value['OrderDemand']."',";
- $train_set = $train_set . "['" . $value['Date'] . "'," . $value['OrderDemand'] . "],";
-
-}
-
-var_dump($train_set);
-
-
-
-$curl = curl_init();
-curl_setopt_array($curl, array(
- CURLOPT_URL => 'http://127.0.0.1:5000/result_validate',
- CURLOPT_RETURNTRANSFER => true,
- CURLOPT_ENCODING => '',
- CURLOPT_MAXREDIRS => 10,
- CURLOPT_TIMEOUT => 0,
- CURLOPT_FOLLOWLOCATION => true,
- CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
- CURLOPT_CUSTOMREQUEST => 'GET',
-));
$validate = curl_exec($curl);
@@ -55,17 +24,25 @@ curl_close($curl);
$arr_new = json_decode($validate, true);
+
$date_validate = '';
$order_validate = '';
$predection = '';
+$error = '';
foreach ($arr_new as $key => $value) {
- $dates = $dates."'".$value['Date']."',";
- $date_validate = $date_validate . "'" . $value['Date'] . "',";
- $predection = $predection . "'" . $value['Predictions'] . "',";
+
+ $date_validate = $date_validate . "'" .$value['Date']. "',";
+ $predection = $predection . "'" .round( $value['Predictions']). "',";
$order_validate= $order_validate."'".$value['OrderDemand']."',";
+ $error_val=$value['OrderDemand']-$value['Predictions'];
+ $error = $error."'". round($error_val)."',";
}
+//==============================================================================
+
+// Fetch warehouse details from database
+