]> source.dussan.org Git - poi.git/commitdiff
Bugzilla 54356 - Support of statistical function SLOPE
authorYegor Kozlov <yegor@apache.org>
Sat, 19 Jan 2013 18:33:34 +0000 (18:33 +0000)
committerYegor Kozlov <yegor@apache.org>
Sat, 19 Jan 2013 18:33:34 +0000 (18:33 +0000)
git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1435633 13f79535-47bb-0310-9956-ffa450edef68

src/java/org/apache/poi/ss/formula/eval/FunctionEval.java
src/java/org/apache/poi/ss/formula/functions/Intercept.java
src/java/org/apache/poi/ss/formula/functions/LinearRegressionFunction.java [new file with mode: 0644]
src/java/org/apache/poi/ss/formula/functions/Slope.java [new file with mode: 0644]
src/testcases/org/apache/poi/ss/formula/functions/TestSlope.java [new file with mode: 0644]

index 021d6980f08623c0cef21f378d4559db2d49632f..733b14b26b9fc31e1b6c63118d1cb751a4a146c6 100644 (file)
@@ -28,7 +28,7 @@ import java.util.TreeSet;
 
 /**
  * @author Amol S. Deshmukh &lt; amolweb at ya hoo dot com &gt;
- * @author Johan Karlsteen - added Intercept
+ * @author Johan Karlsteen - added Intercept and Slope
  */
 public final class FunctionEval {
        /**
@@ -210,6 +210,7 @@ public final class FunctionEval {
                retval[305] = new Sumx2py2();
 
                retval[311] = new Intercept();
+               retval[315] = new Slope();
 
                retval[318] = AggregateFunction.DEVSQ;
 
index 06bb6f97db453ed73ab8f0fff871025bb49b4b3c..cf76cc5880cc6bda7870b49ffeeba5390edba614 100644 (file)
 
 package org.apache.poi.ss.formula.functions;
 
-import org.apache.poi.ss.formula.TwoDEval;
-import org.apache.poi.ss.formula.eval.ErrorEval;
-import org.apache.poi.ss.formula.eval.EvaluationException;
-import org.apache.poi.ss.formula.eval.NumberEval;
-import org.apache.poi.ss.formula.eval.RefEval;
 import org.apache.poi.ss.formula.eval.ValueEval;
-import org.apache.poi.ss.formula.functions.LookupUtils.ValueVector;
+import org.apache.poi.ss.formula.functions.LinearRegressionFunction.FUNCTION;
 
 /**
  * Implementation of Excel function INTERCEPT()<p/>
@@ -40,184 +35,15 @@ import org.apache.poi.ss.formula.functions.LookupUtils.ValueVector;
  */
 public final class Intercept extends Fixed2ArgFunction {
 
-       private static abstract class ValueArray implements ValueVector {
-               private final int _size;
-               protected ValueArray(int size) {
-                       _size = size;
-               }
-
-               public ValueEval getItem(int index) {
-                       if (index < 0 || index > _size) {
-                               throw new IllegalArgumentException("Specified index " + index
-                                               + " is outside range (0.." + (_size - 1) + ")");
-                       }
-                       return getItemInternal(index);
-               }
-               protected abstract ValueEval getItemInternal(int index);
-
-               public final int getSize() {
-                       return _size;
-               }
-       }
-
-       private static final class SingleCellValueArray extends ValueArray {
-               private final ValueEval _value;
-               public SingleCellValueArray(ValueEval value) {
-                       super(1);
-                       _value = value;
-               }
-               @Override
-               protected ValueEval getItemInternal(int index) {
-                       return _value;
-               }
-       }
-
-       private static final class RefValueArray extends ValueArray {
-               private final RefEval _ref;
-               public RefValueArray(RefEval ref) {
-                       super(1);
-                       _ref = ref;
-               }
-               @Override
-               protected ValueEval getItemInternal(int index) {
-                       return _ref.getInnerValueEval();
-               }
-       }
-
-       private static final class AreaValueArray extends ValueArray {
-               private final TwoDEval _ae;
-               private final int _width;
-
-               public AreaValueArray(TwoDEval ae) {
-                       super(ae.getWidth() * ae.getHeight());
-                       _ae = ae;
-                       _width = ae.getWidth();
-               }
-               @Override
-               protected ValueEval getItemInternal(int index) {
-                       int rowIx = index / _width;
-                       int colIx = index % _width;
-                       return _ae.getValue(rowIx, colIx);
-               }
+       private final LinearRegressionFunction func;
+       public Intercept() {
+               func = new LinearRegressionFunction(FUNCTION.INTERCEPT);
        }
-
        
+       @Override
        public ValueEval evaluate(int srcRowIndex, int srcColumnIndex,
                        ValueEval arg0, ValueEval arg1) {
-               double result;
-               try {
-                       ValueVector vvX = createValueVector(arg0);
-                       ValueVector vvY = createValueVector(arg1);
-                       int size = vvX.getSize();
-                       if (size == 0 || vvY.getSize() != size) {
-                               return ErrorEval.NA;
-                       }
-                       result = evaluateInternal(vvX, vvY, size);
-               } catch (EvaluationException e) {
-                       return e.getErrorEval();
-               }
-               if (Double.isNaN(result) || Double.isInfinite(result)) {
-                       return ErrorEval.NUM_ERROR;
-               }
-               return new NumberEval(result);
+               return func.evaluate(srcRowIndex, srcColumnIndex, arg0, arg1);
        }
-       
-       private double evaluateInternal(ValueVector x, ValueVector y, int size)
-                       throws EvaluationException {
+}
 
-               // error handling is as if the x is fully evaluated before y
-               ErrorEval firstXerr = null;
-               ErrorEval firstYerr = null;
-               boolean accumlatedSome = false;
-               double result = 0.0;
-        // first pass: read in data, compute xbar and ybar
-        double sumx = 0.0, sumy = 0.0;
-        
-               for (int i = 0; i < size; i++) {
-                       ValueEval vx = x.getItem(i);
-                       ValueEval vy = y.getItem(i);
-                       if (vx instanceof ErrorEval) {
-                               if (firstXerr == null) {
-                                       firstXerr = (ErrorEval) vx;
-                                       continue;
-                               }
-                       }
-                       if (vy instanceof ErrorEval) {
-                               if (firstYerr == null) {
-                                       firstYerr = (ErrorEval) vy;
-                                       continue;
-                               }
-                       }
-                       // only count pairs if both elements are numbers
-                       if (vx instanceof NumberEval && vy instanceof NumberEval) {
-                               accumlatedSome = true;
-                               NumberEval nx = (NumberEval) vx;
-                               NumberEval ny = (NumberEval) vy;
-                               sumx  += nx.getNumberValue();
-                   sumy  += ny.getNumberValue();
-                       } else {
-                               // all other combinations of value types are silently ignored
-                       }
-               }
-               double xbar = sumx / size;
-        double ybar = sumy / size;
-               
-                // second pass: compute summary statistics
-        double xxbar = 0.0, xybar = 0.0;
-        for (int i = 0; i < size; i++) {
-                       ValueEval vx = x.getItem(i);
-                       ValueEval vy = y.getItem(i);
-                       
-                       if (vx instanceof ErrorEval) {
-                               if (firstXerr == null) {
-                                       firstXerr = (ErrorEval) vx;
-                                       continue;
-                               }
-                       }
-                       if (vy instanceof ErrorEval) {
-                               if (firstYerr == null) {
-                                       firstYerr = (ErrorEval) vy;
-                                       continue;
-                               }
-                       }
-                       
-                       // only count pairs if both elements are numbers
-                       if (vx instanceof NumberEval && vy instanceof NumberEval) {
-                               NumberEval nx = (NumberEval) vx;
-                               NumberEval ny = (NumberEval) vy;
-                   xxbar += (nx.getNumberValue() - xbar) * (nx.getNumberValue() - xbar);
-                   xybar += (nx.getNumberValue() - xbar) * (ny.getNumberValue() - ybar);
-                       } else {
-                               // all other combinations of value types are silently ignored
-                       }
-        }
-        double beta1 = xybar / xxbar;
-        double beta0 = ybar - beta1 * xbar;
-               
-               if (firstXerr != null) {
-                       throw new EvaluationException(firstXerr);
-               }
-               if (firstYerr != null) {
-                       throw new EvaluationException(firstYerr);
-               }
-               if (!accumlatedSome) {
-                       throw new EvaluationException(ErrorEval.DIV_ZERO);
-               }
-               
-               result = beta0;
-               return result;
-       }
-
-       private static ValueVector createValueVector(ValueEval arg) throws EvaluationException {
-               if (arg instanceof ErrorEval) {
-                       throw new EvaluationException((ErrorEval) arg);
-               }
-               if (arg instanceof TwoDEval) {
-                       return new AreaValueArray((TwoDEval) arg);
-               }
-               if (arg instanceof RefEval) {
-                       return new RefValueArray((RefEval) arg);
-               }
-               return new SingleCellValueArray(arg);
-       }
-}
\ No newline at end of file
diff --git a/src/java/org/apache/poi/ss/formula/functions/LinearRegressionFunction.java b/src/java/org/apache/poi/ss/formula/functions/LinearRegressionFunction.java
new file mode 100644 (file)
index 0000000..740fd05
--- /dev/null
@@ -0,0 +1,236 @@
+/*
+ *  ====================================================================
+ *    Licensed to the Apache Software Foundation (ASF) under one or more
+ *    contributor license agreements.  See the NOTICE file distributed with
+ *    this work for additional information regarding copyright ownership.
+ *    The ASF licenses this file to You under the Apache License, Version 2.0
+ *    (the "License"); you may not use this file except in compliance with
+ *    the License.  You may obtain a copy of the License at
+ *
+ *        http://www.apache.org/licenses/LICENSE-2.0
+ *
+ *    Unless required by applicable law or agreed to in writing, software
+ *    distributed under the License is distributed on an "AS IS" BASIS,
+ *    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ *    See the License for the specific language governing permissions and
+ *    limitations under the License.
+ * ====================================================================
+ */
+
+package org.apache.poi.ss.formula.functions;
+
+import org.apache.poi.ss.formula.TwoDEval;
+import org.apache.poi.ss.formula.eval.ErrorEval;
+import org.apache.poi.ss.formula.eval.EvaluationException;
+import org.apache.poi.ss.formula.eval.NumberEval;
+import org.apache.poi.ss.formula.eval.RefEval;
+import org.apache.poi.ss.formula.eval.ValueEval;
+import org.apache.poi.ss.formula.functions.LookupUtils.ValueVector;
+
+/**
+ * Base class for linear regression functions.
+ *
+ * Calculates the linear regression line that is used to predict y values from x values<br/>
+ * (http://introcs.cs.princeton.edu/java/97data/LinearRegression.java.html)
+ * <b>Syntax</b>:<br/>
+ * <b>INTERCEPT</b>(<b>arrayX</b>, <b>arrayY</b>)<p/>
+ * or
+ * <b>SLOPE</b>(<b>arrayX</b>, <b>arrayY</b>)<p/>
+ *
+ *
+ * @author Johan Karlsteen
+ */
+public final class LinearRegressionFunction extends Fixed2ArgFunction {
+       
+       private static abstract class ValueArray implements ValueVector {
+               private final int _size;
+               protected ValueArray(int size) {
+                       _size = size;
+               }
+               @Override
+               public ValueEval getItem(int index) {
+                       if (index < 0 || index > _size) {
+                               throw new IllegalArgumentException("Specified index " + index
+                                               + " is outside range (0.." + (_size - 1) + ")");
+                       }
+                       return getItemInternal(index);
+               }
+               protected abstract ValueEval getItemInternal(int index);
+               @Override
+               public final int getSize() {
+                       return _size;
+               }
+       }
+
+       private static final class SingleCellValueArray extends ValueArray {
+               private final ValueEval _value;
+               public SingleCellValueArray(ValueEval value) {
+                       super(1);
+                       _value = value;
+               }
+               @Override
+               protected ValueEval getItemInternal(int index) {
+                       return _value;
+               }
+       }
+
+       private static final class RefValueArray extends ValueArray {
+               private final RefEval _ref;
+               public RefValueArray(RefEval ref) {
+                       super(1);
+                       _ref = ref;
+               }
+               @Override
+               protected ValueEval getItemInternal(int index) {
+                       return _ref.getInnerValueEval();
+               }
+       }
+
+       private static final class AreaValueArray extends ValueArray {
+               private final TwoDEval _ae;
+               private final int _width;
+
+               public AreaValueArray(TwoDEval ae) {
+                       super(ae.getWidth() * ae.getHeight());
+                       _ae = ae;
+                       _width = ae.getWidth();
+               }
+               @Override
+               protected ValueEval getItemInternal(int index) {
+                       int rowIx = index / _width;
+                       int colIx = index % _width;
+                       return _ae.getValue(rowIx, colIx);
+               }
+       }
+
+       public enum FUNCTION {INTERCEPT, SLOPE};
+       public FUNCTION function;
+       
+       public LinearRegressionFunction(FUNCTION function) {
+               this.function = function;
+       }
+       
+       @Override
+       public ValueEval evaluate(int srcRowIndex, int srcColumnIndex,
+                       ValueEval arg0, ValueEval arg1) {
+               double result;
+               try {
+                       ValueVector vvX = createValueVector(arg0);
+                       ValueVector vvY = createValueVector(arg1);
+                       int size = vvX.getSize();
+                       if (size == 0 || vvY.getSize() != size) {
+                               return ErrorEval.NA;
+                       }
+                       result = evaluateInternal(vvX, vvY, size);
+               } catch (EvaluationException e) {
+                       return e.getErrorEval();
+               }
+               if (Double.isNaN(result) || Double.isInfinite(result)) {
+                       return ErrorEval.NUM_ERROR;
+               }
+               return new NumberEval(result);
+       }
+       
+       private double evaluateInternal(ValueVector x, ValueVector y, int size)
+                       throws EvaluationException {
+
+               // error handling is as if the x is fully evaluated before y
+               ErrorEval firstXerr = null;
+               ErrorEval firstYerr = null;
+               boolean accumlatedSome = false;
+               double result = 0.0;
+        // first pass: read in data, compute xbar and ybar
+        double sumx = 0.0, sumy = 0.0;
+        
+               for (int i = 0; i < size; i++) {
+                       ValueEval vx = x.getItem(i);
+                       ValueEval vy = y.getItem(i);
+                       if (vx instanceof ErrorEval) {
+                               if (firstXerr == null) {
+                                       firstXerr = (ErrorEval) vx;
+                                       continue;
+                               }
+                       }
+                       if (vy instanceof ErrorEval) {
+                               if (firstYerr == null) {
+                                       firstYerr = (ErrorEval) vy;
+                                       continue;
+                               }
+                       }
+                       // only count pairs if both elements are numbers
+                       if (vx instanceof NumberEval && vy instanceof NumberEval) {
+                               accumlatedSome = true;
+                               NumberEval nx = (NumberEval) vx;
+                               NumberEval ny = (NumberEval) vy;
+                               sumx  += nx.getNumberValue();
+                   sumy  += ny.getNumberValue();
+                       } else {
+                               // all other combinations of value types are silently ignored
+                       }
+               }
+               double xbar = sumx / size;
+        double ybar = sumy / size;
+               
+                // second pass: compute summary statistics
+        double xxbar = 0.0, xybar = 0.0;
+        for (int i = 0; i < size; i++) {
+                       ValueEval vx = x.getItem(i);
+                       ValueEval vy = y.getItem(i);
+                       
+                       if (vx instanceof ErrorEval) {
+                               if (firstXerr == null) {
+                                       firstXerr = (ErrorEval) vx;
+                                       continue;
+                               }
+                       }
+                       if (vy instanceof ErrorEval) {
+                               if (firstYerr == null) {
+                                       firstYerr = (ErrorEval) vy;
+                                       continue;
+                               }
+                       }
+                       
+                       // only count pairs if both elements are numbers
+                       if (vx instanceof NumberEval && vy instanceof NumberEval) {
+                               NumberEval nx = (NumberEval) vx;
+                               NumberEval ny = (NumberEval) vy;
+                   xxbar += (nx.getNumberValue() - xbar) * (nx.getNumberValue() - xbar);
+                   xybar += (nx.getNumberValue() - xbar) * (ny.getNumberValue() - ybar);
+                       } else {
+                               // all other combinations of value types are silently ignored
+                       }
+        }
+        double beta1 = xybar / xxbar;
+        double beta0 = ybar - beta1 * xbar;
+               
+               if (firstXerr != null) {
+                       throw new EvaluationException(firstXerr);
+               }
+               if (firstYerr != null) {
+                       throw new EvaluationException(firstYerr);
+               }
+               if (!accumlatedSome) {
+                       throw new EvaluationException(ErrorEval.DIV_ZERO);
+               }
+               
+               if(function == FUNCTION.INTERCEPT) {
+                       return beta0;
+               } else {
+                       return beta1;
+               }
+       }
+
+       private static ValueVector createValueVector(ValueEval arg) throws EvaluationException {
+               if (arg instanceof ErrorEval) {
+                       throw new EvaluationException((ErrorEval) arg);
+               }
+               if (arg instanceof TwoDEval) {
+                       return new AreaValueArray((TwoDEval) arg);
+               }
+               if (arg instanceof RefEval) {
+                       return new RefValueArray((RefEval) arg);
+               }
+               return new SingleCellValueArray(arg);
+       }
+}
+
diff --git a/src/java/org/apache/poi/ss/formula/functions/Slope.java b/src/java/org/apache/poi/ss/formula/functions/Slope.java
new file mode 100644 (file)
index 0000000..ec72102
--- /dev/null
@@ -0,0 +1,49 @@
+/*
+ *  ====================================================================
+ *    Licensed to the Apache Software Foundation (ASF) under one or more
+ *    contributor license agreements.  See the NOTICE file distributed with
+ *    this work for additional information regarding copyright ownership.
+ *    The ASF licenses this file to You under the Apache License, Version 2.0
+ *    (the "License"); you may not use this file except in compliance with
+ *    the License.  You may obtain a copy of the License at
+ *
+ *        http://www.apache.org/licenses/LICENSE-2.0
+ *
+ *    Unless required by applicable law or agreed to in writing, software
+ *    distributed under the License is distributed on an "AS IS" BASIS,
+ *    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ *    See the License for the specific language governing permissions and
+ *    limitations under the License.
+ * ====================================================================
+ */
+
+package org.apache.poi.ss.formula.functions;
+
+import org.apache.poi.ss.formula.eval.ValueEval;
+import org.apache.poi.ss.formula.functions.LinearRegressionFunction.FUNCTION;
+
+/**
+ * Implementation of Excel function SLOPE()<p/>
+ *
+ * Calculates the SLOPE of the linear regression line that is used to predict y values from x values<br/>
+ * (http://introcs.cs.princeton.edu/java/97data/LinearRegression.java.html)
+ * <b>Syntax</b>:<br/>
+ * <b>SLOPE</b>(<b>arrayX</b>, <b>arrayY</b>)<p/>
+ *
+ *
+ * @author Johan Karlsteen
+ */
+public final class Slope extends Fixed2ArgFunction {
+       
+       private final LinearRegressionFunction func;
+       public Slope() {
+               func = new LinearRegressionFunction(FUNCTION.SLOPE);
+       }
+       
+       @Override
+       public ValueEval evaluate(int srcRowIndex, int srcColumnIndex,
+                       ValueEval arg0, ValueEval arg1) {
+               return func.evaluate(srcRowIndex, srcColumnIndex, arg0, arg1);
+       }
+}
+
diff --git a/src/testcases/org/apache/poi/ss/formula/functions/TestSlope.java b/src/testcases/org/apache/poi/ss/formula/functions/TestSlope.java
new file mode 100644 (file)
index 0000000..2ea0332
--- /dev/null
@@ -0,0 +1,137 @@
+/*
+ *  ====================================================================
+ *    Licensed to the Apache Software Foundation (ASF) under one or more
+ *    contributor license agreements.  See the NOTICE file distributed with
+ *    this work for additional information regarding copyright ownership.
+ *    The ASF licenses this file to You under the Apache License, Version 2.0
+ *    (the "License"); you may not use this file except in compliance with
+ *    the License.  You may obtain a copy of the License at
+ *
+ *        http://www.apache.org/licenses/LICENSE-2.0
+ *
+ *    Unless required by applicable law or agreed to in writing, software
+ *    distributed under the License is distributed on an "AS IS" BASIS,
+ *    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ *    See the License for the specific language governing permissions and
+ *    limitations under the License.
+ * ====================================================================
+ */
+
+package org.apache.poi.ss.formula.functions;
+
+import junit.framework.TestCase;
+
+import org.apache.poi.ss.formula.eval.ErrorEval;
+import org.apache.poi.ss.formula.eval.NumberEval;
+import org.apache.poi.ss.formula.eval.ValueEval;
+/**
+ * Test for Excel function SLOPE()
+ *
+ * @author Johan Karlsteen
+ */
+public final class TestSlope extends TestCase {
+       private static final Function SLOPE = new Slope();
+
+       private static ValueEval invoke(Function function, ValueEval xArray, ValueEval yArray) {
+               ValueEval[] args = new ValueEval[] { xArray, yArray, };
+               return function.evaluate(args, -1, (short)-1);
+       }
+
+       private void confirm(Function function, ValueEval xArray, ValueEval yArray, double expected) {
+               ValueEval result = invoke(function, xArray, yArray);
+               assertEquals(NumberEval.class, result.getClass());
+               assertEquals(expected, ((NumberEval)result).getNumberValue(), 0);
+       }
+       private void confirmError(Function function, ValueEval xArray, ValueEval yArray, ErrorEval expectedError) {
+               ValueEval result = invoke(function, xArray, yArray);
+               assertEquals(ErrorEval.class, result.getClass());
+               assertEquals(expectedError.getErrorCode(), ((ErrorEval)result).getErrorCode());
+       }
+
+       private void confirmError(ValueEval xArray, ValueEval yArray, ErrorEval expectedError) {
+               confirmError(SLOPE, xArray, yArray, expectedError);
+       }
+
+       public void testBasic() {
+               Double exp = Math.pow(10, 7.5);
+               ValueEval[] xValues = {
+                       new NumberEval(3+exp),
+                       new NumberEval(4+exp),
+                       new NumberEval(2+exp),
+                       new NumberEval(5+exp),
+                       new NumberEval(4+exp),
+                       new NumberEval(7+exp),
+               };
+               ValueEval areaEvalX = createAreaEval(xValues);
+
+               ValueEval[] yValues = {
+                       new NumberEval(1),
+                       new NumberEval(2),
+                       new NumberEval(3),
+                       new NumberEval(4),
+                       new NumberEval(5),
+                       new NumberEval(6),
+               };
+               ValueEval areaEvalY = createAreaEval(yValues);
+               confirm(SLOPE, areaEvalX, areaEvalY, 0.7752808988764045);
+               // Excel 2010 gives 0.775280898876405
+       }
+
+       /**
+        * number of items in array is not limited to 30
+        */
+       public void testLargeArrays() {
+               ValueEval[] xValues = createMockNumberArray(100, 3); // [1,2,0,1,2,0,...,0,1]
+               xValues[0] = new NumberEval(2.0); // Changes first element to 2
+               ValueEval[] yValues = createMockNumberArray(100, 101); // [1,2,3,4,...,99,100]
+
+               confirm(SLOPE, createAreaEval(xValues), createAreaEval(yValues), -1.231527093596059);
+               // Excel 2010 gives -1.23152709359606
+       }
+
+       private ValueEval[] createMockNumberArray(int size, double value) {
+               ValueEval[] result = new ValueEval[size];
+               for (int i = 0; i < result.length; i++) {
+                       result[i] = new NumberEval((i+1)%value);
+               }
+               return result;
+       }
+
+       private static ValueEval createAreaEval(ValueEval[] values) {
+               String refStr = "A1:A" + values.length;
+               return EvalFactory.createAreaEval(refStr, values);
+       }
+
+       public void testErrors() {
+               ValueEval[] xValues = {
+                               ErrorEval.REF_INVALID,
+                               new NumberEval(2),
+               };
+               ValueEval areaEvalX = createAreaEval(xValues);
+               ValueEval[] yValues = {
+                               new NumberEval(2),
+                               ErrorEval.NULL_INTERSECTION,
+               };
+               ValueEval areaEvalY = createAreaEval(yValues);
+               ValueEval[] zValues = { // wrong size
+                               new NumberEval(2),
+               };
+               ValueEval areaEvalZ = createAreaEval(zValues);
+
+               // if either arg is an error, that error propagates
+               confirmError(ErrorEval.REF_INVALID, ErrorEval.NAME_INVALID, ErrorEval.REF_INVALID);
+               confirmError(areaEvalX, ErrorEval.NAME_INVALID, ErrorEval.NAME_INVALID);
+               confirmError(ErrorEval.NAME_INVALID, areaEvalX, ErrorEval.NAME_INVALID);
+
+               // array sizes must match
+               confirmError(areaEvalX, areaEvalZ, ErrorEval.NA);
+               confirmError(areaEvalZ, areaEvalY, ErrorEval.NA);
+
+               // any error in an array item propagates up
+               confirmError(areaEvalX, areaEvalX, ErrorEval.REF_INVALID);
+
+               // search for errors array by array, not pair by pair
+               confirmError(areaEvalX, areaEvalY, ErrorEval.REF_INVALID);
+               confirmError(areaEvalY, areaEvalX, ErrorEval.NULL_INTERSECTION);
+       }
+}