Changeset 17414 for branches/3040_VectorBasedGP/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/PennML

Timestamp:

01/31/20 12:42:47 (5 years ago)

Author:

pfleck

Message:

#3040 Started adding UCI time series regression benchmarks.
Adapted parser (extracted format options & added parsing for double vectors).

File:

• branches/3040_VectorBasedGP/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/PennML/PennMLRegressionInstanceProvider.cs (modified) (3 diffs)

branches/3040_VectorBasedGP/HeuristicLab.Problems.Instances.DataAnalysis/3.3/Regression/PennML/PennMLRegressionInstanceProvider.cs

@@ -22 +22 @@
 using System;
 using System.Collections.Generic;
-using System.Globalization;
 using System.IO;
 using System.IO.Compression;
@@ -57 +56 @@
       using (var instancesZipFile = new ZipArchive(GetType().Assembly.GetManifestResourceStream(instanceArchiveName), ZipArchiveMode.Read)) {
         foreach (var entry in instancesZipFile.Entries) {
-          NumberFormatInfo numberFormat;
-          DateTimeFormatInfo dateFormat;
-          char separator;
-          using (var stream = entry.Open()) {
-            // the method below disposes the stream 
-            TableFileParser.DetermineFileFormat(stream, out numberFormat, out dateFormat, out separator);
-          }
+          var formatOptions = GetFormatOptions(entry);
 
           using (var stream = entry.Open()) {
@@ -70 +63 @@
 
               // by convention each dataset from the PennML collection reserves the last column for the target
-              var variableNames = header.Split(separator);
+              var variableNames = header.Split(formatOptions.ColumnSeparator);
               var allowedInputVariables = variableNames.Take(variableNames.Length - 1);
               var target = variableNames.Last();