Restart processing is designed to allow users to either restart an interrupted execution of Data Explorer or to add more tables to a completed execution. With certain limitations, it may even be used to add more functions to the analysis. In order to restart an interrupted execution, simply request the Restart option on the OUTPUT tab, along with the same parameters specified for the interrupted execution. Result tables that are not present are created as needed. Any processing that completed is preserved and processing that did not complete is performed newly based on the content of the result tables present. That is, for each function, processing is repeated only for those columns without answers. An exception to this is restricted frequency processing which is repeated in whole for a given table or not at all. See Data Explorer - Frequency Analysis for more information.
In order to add to a previous execution of Data Explorer, the same functions and options should be requested except that new tables (or columns from new tables) should either be added to the request or replace the old tables entirely. If, however, a new function is also requested (as described in Data Explorer - INPUT - Analysis Parameters), the previously selected tables should be included in the request to ensure that the newly requested function is also applied to the originally requested tables. Note that adding new columns to a previously requested table request, or replacing old columns, may lead to errors or unexpected results.
As mentioned previously, if a function such as Statistics, Frequency or Histogram was not requested in the previous run, it may be added with any options that are specific to it. For example, if the previous run successfully performed Values processing only, the other functions may be added to the renewed request for Values processing in the continuation run. In this example, the program will determine that the results for Values processing are already present, so they will be preserved and not rebuilt.
The Data Explorer analysis is parameterized by specifying the tables and/or columns to analyze as well as the desired analysis parameters, expert options, expert where options and OUTPUT options.