static void Main(string[] args)
{
MLContext context = new MLContext();
IDataView dataView = context.Data.LoadFromTextFile <ModelInput>(path: "data.csv", hasHeader: false, separatorChar: ';');
IDataView trainingYearData = context.Data.FilterRowsByColumn(dataView, "Year", upperBound: 2019);
IDataView evaluateYearData = context.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 2020);
var forecastingPipeline = context.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedIncidents",
inputColumnName: "TotalIncidents",
windowSize: 2,
seriesLength: 7,
trainSize: 365,
horizon: 7,
confidenceLevel: 0.5f,
confidenceLowerBoundColumn: "LowerBoundIncidents",
confidenceUpperBoundColumn: "UpperBoundIncidents");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(trainingYearData);
Evaluate(evaluateYearData, forecaster, context);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(context);
forecastEngine.CheckPoint(context, "ML.IncidentAmountForecast.zip");
Forecast(evaluateYearData, 7, forecastEngine, context);
}
static void Main(string[] args)
{
Console.WriteLine("Training time series analysis");
//Step 1. Create a ML Context
var ctx = new MLContext();
string connectionString = "Data Source=localhost;Initial Catalog=kaggle_wallmart;Provider=SQLNCLI11.1;Integrated Security=SSPI;Auto Translate=False;";
connectionString = "Server=localhost;Database=kaggle_wallmart;Integrated Security=True";
string Query = @"
SELECT
CAST(X.[Value] AS REAL) AS [TotalSales],
CAST(Y.date AS DATE) AS [SalesDate],
CAST(year(Y.date) AS REAL) As [Year]
FROM [dbo].[RAW_Train_Eval] AS X
INNER JOIN [dbo].RAW_Calendar AS Y ON Y.d=X.dCode
where Id='HOBBIES_1_278_CA_1_evaluation'
order by 2
";
Console.WriteLine("Connecting to the database...");
//dbChecks dbchecks = new dbChecks();
//dbchecks.ExecuteQuery(connectionString, Query);
System.Data.SqlClient.SqlClientFactory newFactory = SqlClientFactory.Instance;
Console.WriteLine("Loading data...");
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance, connectionString, Query);
DatabaseLoader loader = ctx.Data.CreateDatabaseLoader <ModelInput>();
IDataView dataView = loader.Load(dbSource);
Console.WriteLine($"Loaded {dataView.GetRowCount()} rows...");
IDataView trainingData = ctx.Data.FilterRowsByColumn(dataView, "Year", upperBound: 2016);
IDataView ValidationData = ctx.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 2016);
var forecastingPipeline = ctx.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedSales",
inputColumnName: "TotalSales",
windowSize: 7,
seriesLength: 60,
trainSize: 300,
horizon: 30,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundSales",
confidenceUpperBoundColumn: "UpperBoundSales");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(trainingData);
Evaluate(ValidationData, forecaster, ctx);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(ctx);
forecastEngine.CheckPoint(ctx, "c:\\temp\\Model.zip");
forecastEngine.CheckPoint(ctx, "C:\\Temp\\WallMartModels\\evaluation\\Model_HOBBIES_1_278_CA_1_evaluation.zip");
Forecast(ValidationData, 7, forecastEngine, ctx);
Console.WriteLine("Training time series analysis completed");
}
static void Main(string[] args)
{
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../"));
//string dbFilePath = Path.Combine(rootDir, "Data", "DailyDemand.mdf");
string modelPath = Path.Combine(rootDir, "MLModel.zip");
// var connectionString = $"Data Source=(LocalDB)\\MSSQLLocalDB;AttachDbFilename={dbFilePath};Integrated Security=True;Connect Timeout=30;";
// you can use TblSmartAr.sql and able to make this table
var connectionString = $"Data Source = (localdb)\\MSSQLLocalDB; Initial Catalog = Ar_Database; Integrated Security = True; Connect Timeout = 30; Encrypt = False; TrustServerCertificate = False; ApplicationIntent = ReadWrite; MultiSubnetFailover = False;";
MLContext mlContext = new MLContext();
DatabaseLoader loader = mlContext.Data.CreateDatabaseLoader <ModelInput>();
string query = "SELECT Invoice_Date, CAST(Due_year as REAL) as Due_year, CAST(OverDue_Days as REAL) as OverDue_Days FROM TblSmartAR";
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance,
connectionString,
query);
DateTime now = DateTime.Today;
int thisyear = (now.Year);
IDataView dataView = loader.Load(dbSource);
IDataView firstYearData = mlContext.Data.FilterRowsByColumn(dataView, "Due_year", upperBound: thisyear);
IDataView secondYearData = mlContext.Data.FilterRowsByColumn(dataView, "Due_year", lowerBound: thisyear);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedDays",
inputColumnName: "OverDue_Days",
windowSize: 80,
seriesLength: 90,
trainSize: 365,
horizon: 80,
confidenceLevel: .90f,
confidenceLowerBoundColumn: "MinimumDays",
confidenceUpperBoundColumn: "MaximumDays");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstYearData);
Evaluate(secondYearData, forecaster, mlContext);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
forecastEngine.CheckPoint(mlContext, modelPath);
//foreach(var j in )
//{
// if()
//}
Forecast(secondYearData, 40, forecastEngine, mlContext);
Console.ReadKey();
}
static void Main(string[] args)
{
MLContext mlContext = new MLContext();
IDataView data1View = mlContext.Data.LoadFromTextFile <ModelInput>(_data1Path, separatorChar: ',', hasHeader: false);
IDataView data2View = mlContext.Data.LoadFromTextFile <ModelInput>(_data2Path, separatorChar: ',', hasHeader: false);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "forecasted_count",
inputColumnName: "count",
windowSize: 7,
seriesLength: 30,
trainSize: 334,
horizon: 7,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "lower_count",
confidenceUpperBoundColumn: "upper_count");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(data1View);
Evaluate(data2View, forecaster, mlContext);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
//forecastEngine.CheckPoint(mlContext, modelPath);
ModelOutput forecast = forecastEngine.Predict();
IEnumerable <string> forecastOutput =
mlContext.Data.CreateEnumerable <ModelInput>(data2View, reuseRowObject: false)
.Take(7)
.Select((ModelInput data, int index) =>
{
string action_date = data.action_time.ToString("yyyy-MM-dd");
float actual_count = data.count;
float lowerEstimate = Math.Max(0, forecast.lower_count[index]);
float estimate = forecast.forecasted_count[index];
float upperEstimate = forecast.upper_count[index];
return($"Date: {action_date}\n" +
$"Actual Count: {actual_count}\n" +
$"Lower Estimate: {lowerEstimate}\n" +
$"Forecast: {estimate}\n" +
$"Upper Estimate: {upperEstimate}\n");
});
Console.WriteLine("Forecast");
Console.WriteLine("---------------------");
foreach (var prediction in forecastOutput)
{
Console.WriteLine(prediction);
}
Console.ReadKey();
}
internal static void Start()
{
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../", "BikeDemandForecasting"));
string dbFilePath = Path.Combine(rootDir, "Data", "DailyDemand.mdf");
string modelPath = Path.Combine(rootDir, "MLModel.zip");
var connectionString = $"Data Source=(LocalDB)\\MSSQLLocalDB;AttachDbFilename={dbFilePath};Integrated Security=True;Connect Timeout=30;";
MLContext mlContext = new MLContext();
string query = "SELECT RentalDate, CAST(Year as REAL) as Year, CAST(TotalRentals as REAL) as TotalRentals FROM Rentals";
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance, connectionString, query);
DatabaseLoader loader = mlContext.Data.CreateDatabaseLoader <ModelInput>();
IDataView dataView = loader.Load(dbSource);
/*
* Набор данных содержит данные за два года. Для обучения используются только данные за первый год,
* а данные за второй год откладываются для сравнения фактических значений с прогнозом, созданным
* моделью. Отфильтруйте данные с помощью преобразования FilterRowsByColumn.
*/
IDataView firstYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", upperBound: 1);
IDataView secondYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 1);
//Определите конвейер, который использует SsaForecastingEstimator для прогнозирования значений в наборе данных временных рядов
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedRentals",
inputColumnName: "TotalRentals",
windowSize: 7,
seriesLength: 30,
trainSize: 365,
horizon: 7,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundRentals",
confidenceUpperBoundColumn: "UpperBoundRentals");
//обучить модель и подогнать данные по ранее определенным forecastingPipeline.
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstYearData);
//Оценка модели
Evaluate(secondYearData, forecaster, mlContext);
//создайте TimeSeriesPredictionEngine. TimeSeriesPredictionEngine — это удобный метод для создания единичных прогнозов
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
//Сохраните модель в файл с именем MLModel.zip, которое задано ранее определенной переменной
//modelPath. Выполните метод Checkpoint, чтобы сохранить модель.
forecastEngine.CheckPoint(mlContext, modelPath);
Forecast(secondYearData, 7, forecastEngine, mlContext);
}
static void Main(string[] args)
{
#region 定义相关路径
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../"));
string dbFilePath = Path.Combine(rootDir, "Data", "DailyDemand.mdf");
string modelPath = Path.Combine(rootDir, "MLModel.zip");
var connectionString = $"Data Source=(LocalDB)\\MSSQLLocalDB;AttachDbFilename={dbFilePath};Integrated Security=True;Connect Timeout=30;";
#endregion
//创建ML环境
MLContext mlContext = new MLContext();
//加载数据加载器
DatabaseLoader loader = mlContext.Data.CreateDatabaseLoader <ModelInput>();
//建立具有查询语句的数据源
string query = "SELECT RentalDate, CAST(Year as REAL) as Year, CAST(TotalRentals as REAL) as TotalRentals FROM Rentals";
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance,
connectionString,
query);
//使用加载器加载数据源
IDataView dataView = loader.Load(dbSource);
//获取第一年数据，用于训练，<1
IDataView firstYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", upperBound: 1);
//获取第二年数据，用于预测,>=1
IDataView secondYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 1);
//训练模型
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedRentals",
inputColumnName: "TotalRentals",
windowSize: 7, //7条数据为一个时段分析
seriesLength: 30,
trainSize: 365, //获取365条数据
horizon: 7, //预测7个值
confidenceLevel: 0.95f, //预测的值得是上下限范围内
confidenceLowerBoundColumn: "LowerBoundRentals",
confidenceUpperBoundColumn: "UpperBoundRentals");
//开始训练
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstYearData);
//进行评估
Evaluate(secondYearData, forecaster, mlContext);
//保存训练之后的模型，以用于其他项目
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
forecastEngine.CheckPoint(mlContext, modelPath);
//进行预测
Forecast(secondYearData, 7, forecastEngine, mlContext);
Console.WriteLine("Hello World!");
}
static void Main(string[] args)
{
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../"));
string dbFilePath = Path.Combine(rootDir, "Data", "DailyDemand.mdf");
string modelPath = Path.Combine(rootDir, "MLModel.zip");
var connectionString = $"Data Source=(LocalDB)\\MSSQLLocalDB;AttachDbFilename={dbFilePath};Integrated Security=True;Connect Timeout=30;";
MLContext mlContext = new MLContext();
DatabaseLoader loader = mlContext.Data.CreateDatabaseLoader <ModelInput>();
string query = "SELECT RentalDate, CAST(Year as REAL) as Year, CAST(TotalRentals as REAL) as TotalRentals FROM Rentals";
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance,
connectionString,
query);
IDataView dataView = loader.Load(dbSource);
IDataView firstYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", upperBound: 1);
IDataView secondYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 1);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedRentals",
inputColumnName: "TotalRentals",
windowSize: 7,
seriesLength: 30,
trainSize: 365,
horizon: 7,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundRentals",
confidenceUpperBoundColumn: "UpperBoundRentals");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstYearData);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
forecastEngine.CheckPoint(mlContext, modelPath);
Forecast(secondYearData, 7, forecastEngine, mlContext);
Console.ReadKey();
}
static void Main(string[] args)
{
Console.WriteLine("Hello World!");
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../"));
string dbFilePath = Path.Combine(rootDir, "Data", "DailyDemand.mdf");
string modelPath = Path.Combine(rootDir, "MLModel.zip");
var connectionString = $"Server=localhost,1433;Database=master;User Id=sa;Password=Toffalini1991*;";
MLContext mlContext = new MLContext();
DatabaseLoader loader = mlContext.Data.CreateDatabaseLoader <ModelInput>();
string query = "SELECT RentalDate, CAST(Year as REAL) as Year, CAST(TotalRentals as REAL) as TotalRentals FROM Rentals";
DatabaseSource dbSource = new DatabaseSource(SqlClientFactory.Instance,
connectionString,
query);
IDataView dataView = loader.Load(dbSource);
IDataView firstYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", upperBound: 1);
IDataView secondYearData = mlContext.Data.FilterRowsByColumn(dataView, "Year", lowerBound: 1);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedRentals",
inputColumnName: "TotalRentals",
windowSize: 7,
seriesLength: 30,
trainSize: 365,
horizon: 7,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundRentals",
confidenceUpperBoundColumn: "UpperBoundRentals");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstYearData);
Evaluate(secondYearData, forecaster, mlContext);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
forecastEngine.CheckPoint(mlContext, modelPath);
Forecast(secondYearData, 7, forecastEngine, mlContext);
}
static void Main(string[] args)
{
MLContext _mlContext = new MLContext();
string ConnectionString = "Data Source=146.59.229.11;Initial Catalog=Workshop;User ID=admin;Password=EPSIworkshop2020*";
string rootDir = Path.GetFullPath(Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "../../../"));
string modelPath = Path.Combine(rootDir, "Data", "MLModel.zip");
string query = "SELECT f.matricule AS Feu, CAST(e.nbPassant as REAL) AS NbPassants, CAST(e.jour as REAL) AS Jour, CAST(e.numWeek as REAL) as Semaine " +
"FROM etat e " +
"INNER JOIN feu f ON e.idFeu = f.idFeu " +
"WHERE e.idFeu = 17 AND e.numWeek = 47";
DatabaseLoader loader = _mlContext.Data.CreateDatabaseLoader <ModelInput>();
DatabaseSource databaseSource = new DatabaseSource(SqlClientFactory.Instance, ConnectionString, query);
IDataView dataView = loader.Load(databaseSource);
IDataView firstWeekData = _mlContext.Data.FilterRowsByColumn(dataView, "Jour", upperBound: 1);
IDataView nextWeekData = _mlContext.Data.FilterRowsByColumn(dataView, "Jour", lowerBound: 1);
var forecastingPipeline = _mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedPassants",
inputColumnName: "NbPassants",
windowSize: 7,
seriesLength: 24,
trainSize: 10079,
horizon: 7,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundPassants",
confidenceUpperBoundColumn: "UpperBoundPassants");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(firstWeekData);
Evaluate(nextWeekData, forecaster, _mlContext);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(_mlContext);
forecastEngine.CheckPoint(_mlContext, modelPath);
Forecast(nextWeekData, 4, forecastEngine, _mlContext);
}
/// <summary>
///
/// </summary>
/// <param name="value">Close,High,Low,Open</param>
public static ModelOutput CreateModel(string value)
{
// Load Data
IDataView trainingDataView = mlContext.Data.LoadFromTextFile <ModelInput>(
path: TRAIN_DATA_FILEPATH,
hasHeader: true,
separatorChar: ',',
allowQuoting: true,
allowSparse: false);
var trainList = mlContext.Data.CreateEnumerable <ModelInput>(trainingDataView, true);
int trainSize = trainList.Count();
var testList = trainList.TakeLast(60);
IDataView testDataView = mlContext.Data.LoadFromEnumerable(testList);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedValue",
inputColumnName: value,
windowSize: 5,
seriesLength: 60,
trainSize: trainSize,
horizon: 1,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundValue",
confidenceUpperBoundColumn: "UpperBoundValue");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(trainingDataView);
Evaluate(testDataView, forecaster, mlContext, value);
var forecastEngine = forecaster.CreateTimeSeriesEngine <ModelInput, ModelOutput>(mlContext);
forecastEngine.CheckPoint(mlContext, MODEL_FILE);
return(Forecast(testList, 1, forecastEngine, mlContext, value));
}
private static void RunModel()
{
var modelRunner = new ModelRunner();
var dataSource = JsonConvert.DeserializeObject <List <ModelRunner.ModelInput> >(File.ReadAllText(DataFileName));
string modelPath = Path.Combine(ModelFileName);
MLContext mlContext = new MLContext();
IDataView dataView = mlContext.Data.LoadFromEnumerable(dataSource);
IDataView AllData = mlContext.Data.FilterRowsByColumn(dataView, "YearInSeries", 0, 3);
IDataView Year2017Data = mlContext.Data.FilterRowsByColumn(dataView, "YearInSeries", 2, 3);
var forecastingPipeline = mlContext.Forecasting.ForecastBySsa(
outputColumnName: "ForecastedTotalSold",
inputColumnName: "TotalSold",
windowSize: 30,
seriesLength: 365,
trainSize: 1095,
horizon: 365,
confidenceLevel: 0.95f,
confidenceLowerBoundColumn: "LowerBoundTotalSold",
confidenceUpperBoundColumn: "UpperBoundTotalSold");
SsaForecastingTransformer forecaster = forecastingPipeline.Fit(AllData);
modelRunner.Evaluate(Year2017Data, forecaster, mlContext);
var forecastingEngine = forecaster.CreateTimeSeriesEngine <ModelRunner.ModelInput, ModelRunner.ModelOutput>(mlContext);
forecastingEngine.CheckPoint(mlContext, modelPath);
modelRunner.Forecast(Year2017Data, 365, forecastingEngine, mlContext);
}
