kevingurney commented on code in PR #37627:
URL: https://github.com/apache/arrow/pull/37627#discussion_r1319937008
##########
matlab/src/matlab/+arrow/+tabular/RecordBatch.m:
##########
@@ -91,6 +93,46 @@
function T = toMATLAB(obj)
T = obj.table();
end
+
+ function tf = isequal(obj, varargin)
+ narginchk(2, inf);
+ tf = false;
+
+ schemasToCompare = cell([1 numel(varargin)]);
+ for ii = 1:numel(varargin)
+ rb = varargin{ii};
+ if ~isa(rb, "arrow.tabular.RecordBatch")
+ % If rb is not a RecordBatch, than it cannot be equal
Review Comment:
than -> then
##########
matlab/src/matlab/+arrow/+tabular/RecordBatch.m:
##########
@@ -91,6 +93,46 @@
function T = toMATLAB(obj)
T = obj.table();
end
+
+ function tf = isequal(obj, varargin)
+ narginchk(2, inf);
+ tf = false;
+
+ schemasToCompare = cell([1 numel(varargin)]);
+ for ii = 1:numel(varargin)
+ rb = varargin{ii};
+ if ~isa(rb, "arrow.tabular.RecordBatch")
+ % If rb is not a RecordBatch, than it cannot be equal
+ % to obj. Return false early.
+ return;
+ end
+ schemasToCompare{ii} = rb.Schema;
+ end
+
+ if ~isequal(obj.Schema, schemasToCompare{:})
+ % If the schemas are not equal, the record batches are not
+ % equal. Return false early.
+ return;
+ end
+
+ % Function that extracts the column stored at colIndex from the
+ % record batch stored at rbIndex in varargin.
+ getColulmnFcn = @(rbIndex, colIndex)
varargin{rbIndex}.column(colIndex);
Review Comment:
`getColulmnFcn` -> `getColumnFcn`
##########
matlab/src/matlab/+arrow/+tabular/RecordBatch.m:
##########
@@ -91,6 +93,46 @@
function T = toMATLAB(obj)
T = obj.table();
end
+
+ function tf = isequal(obj, varargin)
+ narginchk(2, inf);
+ tf = false;
+
+ schemasToCompare = cell([1 numel(varargin)]);
+ for ii = 1:numel(varargin)
+ rb = varargin{ii};
+ if ~isa(rb, "arrow.tabular.RecordBatch")
+ % If rb is not a RecordBatch, than it cannot be equal
+ % to obj. Return false early.
+ return;
+ end
+ schemasToCompare{ii} = rb.Schema;
+ end
+
+ if ~isequal(obj.Schema, schemasToCompare{:})
+ % If the schemas are not equal, the record batches are not
+ % equal. Return false early.
+ return;
+ end
+
+ % Function that extracts the column stored at colIndex from the
+ % record batch stored at rbIndex in varargin.
+ getColulmnFcn = @(rbIndex, colIndex)
varargin{rbIndex}.column(colIndex);
+
+ rbIndices = 1:numel(varargin);
+ for ii = 1:obj.NumColumns
+ colIndices = repmat(ii, [1 numel(rbIndices)]);
+ % Gather all columns at index ii across within the record
Review Comment:
across within -> across
##########
matlab/test/arrow/tabular/tRecordBatch.m:
##########
@@ -386,6 +386,78 @@ function ErrorIfColumnNameIsNonScalar(testCase)
testCase.verifyError(@() recordBatch.column(name),
"arrow:badsubscript:NonScalar");
end
+ function TestIsEqualTrue(testCase)
+ % Verify two record batches are considered equal if:
+ % 1. They have the same schema
+ % 2. Their corresponding columns are equal
+ import arrow.tabular.RecordBatch
+
+ a1 = arrow.array([1 2 3]);
+ a2 = arrow.array(["A" "B" "C"]);
+ a3 = arrow.array([true true false]);
+
+ rb1 = RecordBatch.fromArrays(a1, a2, a3, ...
+ ColumnNames=["A", "B", "C"]);
+ rb2 = RecordBatch.fromArrays(a1, a2, a3, ...
+ ColumnNames=["A", "B", "C"]);
+ testCase.verifyTrue(isequal(rb1, rb2));
+
+ % Compare zero-column record batches
+ rb3 = RecordBatch.fromArrays();
+ rb4 = RecordBatch.fromArrays();
+ testCase.verifyTrue(isequal(rb3, rb4));
+
+ % Compare zero-row record batches
+ a4 = arrow.array([]);
+ a5 = arrow.array(strings(0, 0));
+ rb5 = RecordBatch.fromArrays(a4, a5, ColumnNames=["D" "E"]);
+ rb6 = RecordBatch.fromArrays(a4, a5, ColumnNames=["D" "E"]);
+ testCase.verifyTrue(isequal(rb5, rb6));
+
+ % Call isequal with more than two arguments
+ testCase.verifyTrue(isequal(rb3, rb4, rb3, rb4));
+ end
+
+ function TestIsEqualFalse(testCase)
+ % Verify isequal returns false when expected.
+ import arrow.tabular.RecordBatch
+
+ a1 = arrow.array([1 2 3]);
+ a2 = arrow.array(["A" "B" "C"]);
+ a3 = arrow.array([true true false]);
+ a4 = arrow.array(["A" missing "C"]);
+ a5 = arrow.array([1 2]);
+ a6 = arrow.array(["A" "B"]);
+ a7 = arrow.array([true true]);
+
+ rb1 = RecordBatch.fromArrays(a1, a2, a3, ...
+ ColumnNames=["A", "B", "C"]);
+ rb2 = RecordBatch.fromArrays(a1, a2, a3, ...
+ ColumnNames=["D", "E", "F"]);
+ rb3 = RecordBatch.fromArrays(a1, a4, a3, ...
+ ColumnNames=["A", "B", "C"]);
+ rb4 = RecordBatch.fromArrays(a5, a6, a7, ...
+ ColumnNames=["A", "B", "C"]);
+ rb5 = RecordBatch.fromArrays(a1, a2, a3, a1, ...
+ ColumnNames=["A", "B", "C", "D"]);
+
+ % The column names are not equal
+ testCase.verifyFalse(isequal(rb1, rb2));
+
+ % The columns are not equal
+ testCase.verifyFalse(isequal(rb1, rb3));
+
+ % The number of rows are not equal
+ testCase.verifyFalse(isequal(rb1, rb4));
+
+ % The number of columns are not equal
Review Comment:
Indentation is off here.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
