/// <summary>
/// Application entry point.
/// Opens USDA Nutrient database (revision 26), denormalizes it using UsdaDenormalizer library
/// and writes result to a CSV file with a specified name.
/// </summary>
/// <param name="args">
/// [0] path to the directory containing ASCII version of database
/// [1] path to the output CSV file
/// </param>
static void Main(string[] args)
{
// check if input is valid
if (args.Length > 1)
{
string inputDirectoryPath = args[0];
string outputCsvPath = args[1];
if (System.IO.Directory.Exists(inputDirectoryPath))
{
// read nutrient data from files in filesystem to object model in memory
DirectoryDataProvider dataProvider = new DirectoryDataProvider(inputDirectoryPath, 26, new WindowsTextReader());
DatabaseRepresentation dbModel = new DatabaseRepresentation(dataProvider);
// create denormalized (flat) model based using database data
DenormalizedModel denormalizedModel = new DenormalizedModel(dbModel);
// write flat model into CSV file as one big table
using (var writer = System.IO.File.CreateText(outputCsvPath))
{
CsvHelper.CsvWriter csvWriter = new CsvHelper.CsvWriter(writer);
csvWriter.WriteRecords(denormalizedModel.FoodItems);
}
}
else
{
throw new ArgumentException("Specified directory does not exist or you don't have an access to it: " + inputDirectoryPath);
}
}
else
{
throw new ArgumentException("No arguments were provided.");
}
}
/// <summary>
/// Application entry point.
/// Opens USDA Nutrient database (revision 26), denormalizes it using UsdaDenormalizer library
/// and writes result to a CSV file with a specified name.
/// </summary>
/// <param name="args">
/// [0] path to the directory containing ASCII version of database
/// [1] path to the output CSV file
/// </param>
private static void Main(string[] args)
{
// check if input is valid
if (args.Length > 1)
{
var inputDirectoryPath = args[0];
var outputCsvPath = args[1];
if (Directory.Exists(inputDirectoryPath))
{
// read nutrient data from files in filesystem to object model in memory
var dataProvider = new DirectoryDataProvider(inputDirectoryPath, 26, new WindowsTextReader());
var dbModel = new DatabaseRepresentation(dataProvider);
// create denormalized (flat) model based using database data
var denormalizedModel = new DenormalizedModel(dbModel);
// write flat model into CSV file as one big table
using (var writer = File.CreateText(outputCsvPath))
{
var csvWriter = new CsvWriter(writer);
csvWriter.WriteRecords(denormalizedModel.FoodItems);
}
}
else
{
throw new ArgumentException(
"Specified directory does not exist or you don't have an access to it: " + inputDirectoryPath);
}
}
else
{
throw new ArgumentException("No arguments were provided.");
}
}
private PointsOfInterestClassifier Learn(string sourcePath, string truthPath)
{
this.WriteLine(null, "Learning...");
PointsOfInterestClassifier classifier = new PointsOfInterestClassifier();
using (DirectoryDataProvider dataProvider = new DirectoryDataProvider())
{
dataProvider.Add(sourcePath, false, truthPath, "#Class");
ClassifierProgress <TestImage> progress = new ClassifierProgress <TestImage>(
(source, index) =>
{
this.StopwatchRestart();
},
(source, index, answer, exception) =>
{
long duration = this.StopwatchStop();
Interlocked.Increment(ref this.totalImages);
this.Write(
null,
"({0})\tFile: {1} ... ",
index,
source.SourceId.ToFileName(false));
if (exception != null)
{
this.WriteLine(null, "ERROR.");
this.WriteException(null, exception);
}
else
{
this.WriteLine(null, "OK ({0} ms)", duration);
}
});
classifier.Train <TestImage>(
dataProvider.Generate(null),
(x, cancellationToken) => new ImageSource(x.SourceId, x.Image),
x => string.Concat(x.Labels),
progress,
CancellationToken.None);
}
return(classifier);
}
private void Test(PointsOfInterestClassifier classifier, string sourcePath, string truthPath)
{
this.WriteLine(null, "Testing...");
List <ClassificationResult <string> > results = new List <ClassificationResult <string> >();
using (DirectoryDataProvider dataProvider = new DirectoryDataProvider())
{
dataProvider.Add(sourcePath, false, truthPath, "#Class");
ClassifierProgress <TestImage> progress = new ClassifierProgress <TestImage>(
(source, index) =>
{
this.StopwatchRestart();
},
(source, index, answer, exception) =>
{
long duration = this.StopwatchStop();
Interlocked.Increment(ref this.totalImages);
this.Write(
null,
"({0})\tFile: {1} ... ",
index,
source.SourceId.ToFileName(false));
if (exception != null)
{
this.WriteLine(null, "ERROR.");
this.WriteException(null, exception);
}
else
{
this.WriteLine(
null,
"OK ({0} ms) {1} {2:F4}",
duration,
answer?.ClassName ?? string.Empty,
answer?.Confidence ?? 0.0);
}
});
foreach ((TestImage image, Answer answer) in classifier.Classify <TestImage>(
dataProvider.Generate(null),
(x, cancellationToken) => new ImageSource(x.SourceId, x.Image),
progress,
CancellationToken.None))
{
results.Add(new ClassificationResult <string>(
answer.Id,
answer.ClassName,
string.Concat(image.Labels),
answer.Confidence,
answer.Confidence >= 0.2f));
}
}
// write report
ClassificationReport <string> testReport = new ClassificationReport <string>(results);
using (StreamWriter outputFile = new StreamWriter(Console.OpenStandardOutput()))
////using (StreamWriter outputFile = File.CreateText(this.options.OutputFileName))
{
ClassificationReportWriter <string> .WriteReport(outputFile, testReport, ClassificationReportMode.All);
}
}
