Aggregate storage data load properties that determine how missing and zero values, duplicate values, and multiple values for the same cell in the data source are processed.
ignore_missing_values: Ignore missing values in the data source.
ignore_zero_values: Ignore zeros in the data source.
aggregate_use_last: Combine duplicate cells by using the value of the cell that was loaded last into the data load buffer. When using this option, data loads are significantly slower, even if there are not any duplicate values.
Caution! | The aggregate_use_last method has significant performance impact, and is not intended for large data loads. If your data load is larger than one million cells, consider separating the numeric data into a separate data load process (from any typed measure data). The separate data load can use aggregate_sum instead. |
aggregate_sum: (Default) Add values when the buffer contains multiple values for the same cell.
If you use multiple properties and any conflict occurs, the last property listed takes precedence.
string (see MaxL Syntax Notes)
alter database (aggregate storage)