Changeset 273 for trunk/sources/HeuristicLab.DataAnalysis

Timestamp:

05/29/08 18:05:50 (17 years ago)

Author:

gkronber

Message:

fixed #160

Location:

trunk/sources/HeuristicLab.DataAnalysis

Files:

• DataFormatException.cs (added)
• DatasetParser.cs (modified) (12 diffs)
• HeuristicLab.DataAnalysis.csproj (modified) (1 diff)

trunk/sources/HeuristicLab.DataAnalysis/DatasetParser.cs

@@ -28 +28 @@
 namespace HeuristicLab.DataAnalysis {
   public class DatasetParser {
+    private const string PROBLEMNAME = "PROBLEMNAME";
+    private const string VARIABLENAMES = "VARIABLENAMES";
+    private const string TARGETVARIABLE = "TARGETVARIABLE";
+    private const string MAXIMUMTREEHEIGHT = "MAXIMUMTREEHEIGHT";
+    private const string MAXIMUMTREESIZE = "MAXIMUMTREESIZE";
+    private const string TRAININGSAMPLESSTART = "TRAININGSAMPLESSTART";
+    private const string TRAININGSAMPLESEND = "TRAININGSAMPLESEND";
     private Tokenizer tokenizer;
     private Dictionary<string, List<Token>> metadata;
@@ -53 +60 @@
     public string ProblemName {
       get {
-        return metadata["PROBLEMNAME"][0].stringValue;
+        if(metadata.ContainsKey(PROBLEMNAME)) {
+          return metadata[PROBLEMNAME][0].stringValue;
+        } else return "-";
       }
     }
@@ -59 +68 @@
     public string[] VariableNames {
       get {
-        List<Token> nameList = metadata["VARIABLENAMES"];
-        string[] names = new string[nameList.Count];
-        for(int i = 0; i < names.Length; i++) {
-          names[i] = nameList[i].stringValue;
-        }
-
-        return names;
+        if(metadata.ContainsKey(VARIABLENAMES)) {
+          List<Token> nameList = metadata[VARIABLENAMES];
+          string[] names = new string[nameList.Count];
+          for(int i = 0; i < names.Length; i++) {
+            names[i] = nameList[i].stringValue;
+          }
+          return names;
+        } else {
+          string[] names = new string[columns];
+          for(int i = 0; i < names.Length; i++) {
+            names[i] = "X" + i.ToString("000");
+          }
+          return names;
+        }
       }
     }
@@ -71 +87 @@
     public int TargetVariable {
       get {
-        return metadata["TARGETVARIABLE"][0].intValue;
+        if(metadata.ContainsKey(TARGETVARIABLE)) {
+          return metadata[TARGETVARIABLE][0].intValue;
+        } else return 0; // default is the first column
       }
     }
@@ -77 +95 @@
     public int MaxTreeHeight {
       get {
-        return metadata["MAXIMUMTREEHEIGHT"][0].intValue;
+        if(metadata.ContainsKey(MAXIMUMTREEHEIGHT)) {
+          return metadata[MAXIMUMTREEHEIGHT][0].intValue;
+        } else return 0;
       }
     }
@@ -83 +103 @@
     public int MaxTreeSize {
       get {
-        return metadata["MAXIMUMTREESIZE"][0].intValue;
+        if(metadata.ContainsKey(MAXIMUMTREESIZE)) {
+          return metadata[MAXIMUMTREESIZE][0].intValue;
+        } else return 0;
       }
     }
@@ -89 +111 @@
     public int TrainingSamplesStart {
       get {
-        if(!metadata.ContainsKey("TRAININGSAMPLESSTART")) return 0;
-        else return metadata["TRAININGSAMPLESSTART"][0].intValue;
+        if(metadata.ContainsKey(TRAININGSAMPLESSTART)) {
+          return metadata[TRAININGSAMPLESSTART][0].intValue;
+        } else return 0;
       }
     }
@@ -96 +119 @@
     public int TrainingSamplesEnd {
       get {
-        if(!metadata.ContainsKey("TRAININGSAMPLESEND")) return rows;
-        else return metadata["TRAININGSAMPLESEND"][0].intValue;
+        if(metadata.ContainsKey(TRAININGSAMPLESEND)) {
+          return metadata[TRAININGSAMPLESEND][0].intValue;
+        } else return rows;
       }
     }
@@ -279 +303 @@
             if(samplesList.Count > 0 && samplesList[0].Count != row.Count) {
               Error("The first row of the dataset has " + samplesList[0].Count + " columns." +
-                "\nLine " + tokenizer.CurrentLineNumber + " has " + row.Count + " columns.");
+                "\nLine " + tokenizer.CurrentLineNumber + " has " + row.Count + " columns.", "", tokenizer.CurrentLineNumber);
             }
           } else if(samplesList.Count > 0) {
@@ -301 +325 @@
           // when we are parsing non-strictly we also allow unreadable values inserting NAN instead
           if(strict) {
-            Error("Unkown value " + current + " in line " + tokenizer.CurrentLineNumber +
-              "\n" + tokenizer.CurrentLine);
+            Error("Unexpected token.", current.stringValue, tokenizer.CurrentLineNumber);
           } else {
             row.Add(double.NaN);
@@ -316 +339 @@
         Token nameToken = tokenizer.Next();
         if(nameToken.type != TokenTypeEnum.String)
-          throw new Exception("Expected a variable name; got " + nameToken +
-            "\nLine " + tokenizer.CurrentLineNumber + ": " + tokenizer.CurrentLine);
+          Error("Expected a variable name.", nameToken.stringValue, tokenizer.CurrentLineNumber);
 
         Expect(Tokenizer.AssignmentToken);
@@ -335 +357 @@
       Token actualToken = tokenizer.Next();
       if(actualToken != expectedToken) {
-        Error("Expected: " + expectedToken + " got: " + actualToken +
-          "\nLine " + tokenizer.CurrentLineNumber + ": " + tokenizer.CurrentLine);
-      }
-    }
-
-    private void Error(string message) {
-      throw new Exception("Error while parsing.\n" + message);
+        Error("Expected: " + expectedToken, actualToken.stringValue, tokenizer.CurrentLineNumber);
+      }
+    }
+
+    private void Error(string message, string token, int lineNumber) {
+      throw new DataFormatException("Error while parsing.\n" + message, token, lineNumber);
     }
     #endregion

trunk/sources/HeuristicLab.DataAnalysis/HeuristicLab.DataAnalysis.csproj

@@ -51 +51 @@
   </ItemGroup>
   <ItemGroup>
+    <Compile Include="DataFormatException.cs" />
     <Compile Include="Dataset.cs" />
     <Compile Include="DatasetParser.cs" />