Nicer Dataset object validation error messages

CHANGELOG.md

@@ -1,9 +1,10 @@
-- Unreleased
+- 0.1.1
     - Fixed Image Resizer placeholder image
     - Fixed Filesystem no write permissions on instantiation
-    - Nicer object string representations
-    - Do not terminate empty spatial leaf nodes
+    - Nicer Stringable object string representations
+    - Do not terminate empty Spatial tree leaf nodes
     - Additional Filesystem persister checks
+    - Nicer Dataset object validation error messages
 
 - 0.1.0
     - CV Report Generators now return Report objects

docs/datasets/labeled.md

@@ -3,7 +3,7 @@
 # Labeled
 A Labeled dataset is used to train supervised learners and for testing a model by providing the ground-truth. In addition to the standard dataset API, a labeled dataset can perform operations such as stratification and sorting the dataset using the label column.
 
-**Note:** Since PHP silently converts integer strings (ex. `'1'`) to integers in some circumstances, you should not use integer strings as class labels. Instead, use an appropriate non-integer string class label such as `first` or `class 1`.
+**Note:** Since PHP silently converts integer strings (ex. `'1'`) to integers in some circumstances, you should not use integer strings as class labels. Instead, use an appropriate non-integer string class name such as `'class 1'`, `'#1'`, or `'first'`.
 
 ## Parameters
 | # | Param | Default | Type | Description |

src/Datasets/Dataset.php

@@ -90,15 +90,15 @@ public function __construct(array $samples = [], bool $validate = true)
                 if (count($sample) !== $n) {
                     throw new InvalidArgumentException('Number of columns'
                         . " must be equal for all samples, $n expected but"
-                        . ' ' . count($sample) . ' given.');
+                        . ' ' . count($sample) . " given at row $row.");
                 }
 
                 foreach ($sample as $column => $value) {
                     if (DataType::detect($value) != $types[$column]) {
                         throw new InvalidArgumentException("Column $column must"
                             . ' contain values of the same data type,'
                             . " $types[$column] expected but "
-                            . DataType::detect($value) . ' given.');
+                            . DataType::detect($value) . " given at row $row.");
                     }
                 }
             }

src/Datasets/Labeled.php

@@ -18,6 +18,10 @@
 use function get_class;
 use function gettype;
 use function array_slice;
+use function is_string;
+use function is_numeric;
+use function is_float;
+use function is_nan;
 
 use const Rubix\ML\PHI;
 use const Rubix\ML\EPSILON;
@@ -148,16 +152,16 @@ public function __construct(array $samples = [], array $labels = [], bool $valid
                     . " categorical or continuous, $type given.");
             }
 
-            foreach ($labels as $label) {
+            foreach ($labels as $offset => $label) {
                 if (DataType::detect($label) != $type) {
-                    throw new InvalidArgumentException('Labels must be'
-                        . " the same data type, $type expected but "
+                    throw new InvalidArgumentException('Invalid label type'
+                        . " found at offset $offset, $type expected but "
                         . DataType::detect($label) . ' given.');
                 }
 
                 if (is_float($label) and is_nan($label)) {
                     throw new InvalidArgumentException('Labels must not'
-                        . ' contain NaN values.');
+                        . " contain NaN values, NaN found at offset $offset.");
                 }
             }
         }