public IActionResult GetProductUnitDemandEstimation(float productId,
[FromQuery] int year, [FromQuery] int month,
[FromQuery] float units, [FromQuery] float avg,
[FromQuery] int count, [FromQuery] float max,
[FromQuery] float min, [FromQuery] float prev)
{
// Build product sample
var inputExample = new ProductData()
{
productId = productId,
year = year,
month = month,
units = units,
avg = avg,
count = count,
max = max,
min = min,
prev = prev
};
ProductUnitRegressionPrediction nextMonthUnitDemandEstimation = null;
//Predict
nextMonthUnitDemandEstimation = this.productSalesModel.Predict(inputExample);
return(Ok(nextMonthUnitDemandEstimation.Score));
}
private static void TestPredictionRegression(MLContext mlContext, string outputModelPath, string dataPath)
{
Console.WriteLine("Testing Product Unit Sales Forecast Regression model");
// Read the model that has been previously saved by the method SaveModel.
ITransformer trainedModel;
using (var stream = File.OpenRead(outputModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
var predictionEngine = mlContext.Model.CreatePredictionEngine <SaleData, ProductUnitRegressionPrediction>(trainedModel);
Console.WriteLine("** Testing Product **");
// Predict the nextperiod/month forecast to the one provided
ProductUnitRegressionPrediction prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[0]);
Console.WriteLine($"Product: {SampleSaleData.MonthlyData[0].id}, month: {SampleSaleData.MonthlyData[0].mes + 1}, year: {SampleSaleData.MonthlyData[0].ano} - Real value (units): {SampleSaleData.MonthlyData[0].siguiente}, Forecast Prediction (units): {prediction.Score}");
// Predicts the nextperiod/month forecast to the one provided
prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[1]);
Console.WriteLine($"Product: {SampleSaleData.MonthlyData[1].id}, month: {SampleSaleData.MonthlyData[1].mes + 1}, year: {SampleSaleData.MonthlyData[1].ano} - Forecast Prediction (units): {prediction.Score}");
//prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[2]);
//Console.WriteLine($"Product: {SampleSaleData.MonthlyData[2].id}, month: {SampleSaleData.MonthlyData[2].mes + 1}, year: {SampleSaleData.MonthlyData[2].ano} - Forecast Prediction (units): {prediction.Score}");
}
/// <summary>
/// Predict samples using saved model
/// </summary>
/// <param name="outputModelPath">Model file path</param>
public static void TestPrediction(MLContext mlContext, string outputModelPath = "product_month_fastTreeTweedie.zip")
{
ConsoleWriteHeader("Testing Product Unit Sales Forecast Regression model");
// Read the model that has been previously saved by the method SaveModel.
ITransformer trainedModel;
using (var stream = File.OpenRead(outputModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
var predictionEngine = mlContext.Model.CreatePredictionEngine <ProductData, ProductUnitRegressionPrediction>(trainedModel);
Console.WriteLine("** Testing Product **");
// Predict the nextperiod/month forecast to the one provided
ProductUnitRegressionPrediction prediction = predictionEngine.Predict(SampleProductData.MonthlyData[0]);
Console.WriteLine($"Product: {SampleProductData.MonthlyData[0].productId}, month: {SampleProductData.MonthlyData[0].month + 1}, year: {SampleProductData.MonthlyData[0].year} - Real value (units): {SampleProductData.MonthlyData[0].next}, Forecast Prediction (units): {prediction.Score}");
// Predicts the nextperiod/month forecast to the one provided
prediction = predictionEngine.Predict(SampleProductData.MonthlyData[1]);
Console.WriteLine($"Product: {SampleProductData.MonthlyData[1].productId}, month: {SampleProductData.MonthlyData[1].month + 1}, year: {SampleProductData.MonthlyData[1].year} - Forecast Prediction (units): {prediction.Score}");
}
private static void TestPredictionRegression(MLContext mlContext, string outputModelPath, string dataPath)
{
Console.WriteLine("Testing Product Unit Sales Forecast Regression model");
// Read the model that has been previously saved by the method SaveModel.
ITransformer trainedModel;
using (var stream = File.OpenRead(outputModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
//using (var reader = new StreamReader(dataPath))
//{
// reader.ReadLine();
// while (!reader.EndOfStream)
// {
// var line = reader.ReadLine();
// var values = line.Split(',');
// int value = Convert.ToInt32(values[4]);
// for (int i = 0; i < value - 1; i++)
// {
// Console.Write("*");
// }
// Console.Write("*\n");
// }
//}
var predictionEngine = mlContext.Model.CreatePredictionEngine <SaleData, ProductUnitRegressionPrediction>(trainedModel);
Console.WriteLine("** Testing Product **");
// Predict the nextperiod/month forecast to the one provided
ProductUnitRegressionPrediction prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[0]);
Console.WriteLine($"Product: {SampleSaleData.MonthlyData[0].id}, month: {SampleSaleData.MonthlyData[0].mes + 1}, year: {SampleSaleData.MonthlyData[0].ano} - Real value (units): {SampleSaleData.MonthlyData[0].siguiente}, Forecast Prediction (units): {prediction.Score}");
// Predicts the nextperiod/month forecast to the one provided
prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[1]);
Console.WriteLine($"Product: {SampleSaleData.MonthlyData[1].id}, month: {SampleSaleData.MonthlyData[1].mes + 1}, year: {SampleSaleData.MonthlyData[1].ano} - Forecast Prediction (units): {prediction.Score}");
//prediction = predictionEngine.Predict(SampleSaleData.MonthlyData[2]);
//Console.WriteLine($"Product: {SampleSaleData.MonthlyData[2].id}, month: {SampleSaleData.MonthlyData[2].mes + 1}, year: {SampleSaleData.MonthlyData[2].ano} - Forecast Prediction (units): {prediction.Score}");
}
/// <summary>
/// Predict samples using saved model
/// </summary>
/// <param name="outputModelPath">Model file path</param>
public static void TestPrediction(MLContext mlContext, string outputModelPath = "product_month_fastTreeTweedie.zip")
{
ConsoleWriteHeader("Testing Product Unit Sales Forecast Regression model");
// Read the model that has been previously saved by the method SaveModel
ITransformer trainedModel;
using (var stream = File.OpenRead(outputModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
var predictionEngine = mlContext.Model.CreatePredictionEngine <ProductData, ProductUnitRegressionPrediction>(trainedModel);
Console.WriteLine("** Testing Product 1 **");
// Build sample data
ProductData dataSample = new ProductData()
{
productId = 263,
month = 10,
year = 2017,
avg = 91,
max = 370,
min = 1,
count = 10,
prev = 1675,
units = 910
};
// Predict the nextperiod/month forecast to the one provided
ProductUnitRegressionPrediction prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Product: {dataSample.productId}, month: {dataSample.month + 1}, year: {dataSample.year} - Real value (units): 551, Forecast Prediction (units): {prediction.Score}");
dataSample = new ProductData()
{
productId = 263,
month = 11,
year = 2017,
avg = 29,
max = 221,
min = 1,
count = 35,
prev = 910,
units = 551
};
// Predicts the nextperiod/month forecast to the one provided
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Product: {dataSample.productId}, month: {dataSample.month + 1}, year: {dataSample.year} - Forecast Prediction (units): {prediction.Score}");
Console.WriteLine(" ");
Console.WriteLine("** Testing Product 2 **");
dataSample = new ProductData()
{
productId = 988,
month = 10,
year = 2017,
avg = 43,
max = 220,
min = 1,
count = 25,
prev = 1036,
units = 1094
};
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Product: {dataSample.productId}, month: {dataSample.month + 1}, year: {dataSample.year} - Real Value (units): 1076, Forecasting (units): {prediction.Score}");
dataSample = new ProductData()
{
productId = 988,
month = 11,
year = 2017,
avg = 41,
max = 225,
min = 4,
count = 26,
prev = 1094,
units = 1076
};
prediction = predictionEngine.Predict(dataSample);
Console.WriteLine($"Product: {dataSample.productId}, month: {dataSample.month + 1}, year: {dataSample.year} - Forecasting (units): {prediction.Score}");
}
public static void PredictionRegression(int id_predict, SqlParameter[] parameters, string outputModelPath = "product_month_fastTreeTweedie.zip")
{
MLContext mlContext = new MLContext(seed: 1);
// Read the model that has been previously saved by the method SaveModel.
ITransformer trainedModel;
using (var stream = File.OpenRead(ModelPath))
{
trainedModel = mlContext.Model.Load(stream, out var modelInputSchema);
}
var predictionEngine = mlContext.Model.CreatePredictionEngine <SaleData, ProductUnitRegressionPrediction>(trainedModel);
DBManagement connect = new DBManagement();
ProductUnitRegressionPrediction prediction = null;
int mescalc = ((DateTime)parameters[3].Value).Month;
int anocalc = ((DateTime)parameters[3].Value).Year;
int anterior = 0;
for (int i = 0; i < (int)parameters[1].Value; i++)
{
SaleData lastMonth = null;
//int mescalc = ((DateTime)parameters[3].Value).Month;
//int anocalc = ((DateTime)parameters[3].Value).Year;
if (i == 0)
{
lastMonth = connect.getLastMonthSaleData(988, anocalc, mescalc);
}
else
{
lastMonth = new SaleData()
{
siguiente = 0,
id = (float)988,
ano = (float)anocalc,
mes = (float)mescalc,
unidades = (float)prediction.Score,
anterior = (float)anterior
};
}
// Predict the nextperiod/month forecast to the one provided
prediction = predictionEngine.Predict(lastMonth);
anterior = (int)lastMonth.unidades;
SqlParameter[] parametersPrevision = connect.getparametersPrevisionArray();
parametersPrevision[0].Value = id_predict;
parametersPrevision[1].Value = lastMonth.id;
if ((int)lastMonth.mes + 1 <= 12)
{
parametersPrevision[2].Value = new DateTime((int)lastMonth.ano, ((int)lastMonth.mes + 1), 1);
}
else
{
parametersPrevision[2].Value = new DateTime((int)lastMonth.ano + 1, 1, 1);
}
parametersPrevision[3].Value = prediction.Score;
connect.insertPrevision(parametersPrevision);
mescalc = (((DateTime)parameters[3].Value).Month + i) % 12 + 1;
anocalc = ((DateTime)parameters[3].Value).Year + (((DateTime)parameters[3].Value).Month + i) / 12;
}
}