Extracting Data

You can extract assumption and forecast information generated during a Crystal Ball simulation. Crystal Ball places the extracted data in the specified worksheet location. You can only extract data after you run a simulation or restore saved results.

  To extract data:

  1. Select Extract Data. Extract Data button, in the Crystal Ball ribbon

  2. In the Extract Data dialog, select the type of data to extract:

    • Statistics — Descriptive statistics summarizing the assumption and forecast values.

    • Percentiles — The probability of achieving values below a particular threshold in the selected increments. Optional: You can reverse the meaning of the percentiles by changing the setting in the Run Preferences, then Options panel (Setting Statistics Preferences).

      Note:

      If you select Percentiles, a dialog opens so you can select which percentiles to use. Optional: Select Custom and enter a set of custom percentiles if the set you need is not already available in the dialog.

    • Chart Bins — For each group interval, or bin, the interval range as well as the probability and frequency of occurrences within the interval for the forecast. This setting is independent of the Chart Preferences density setting that controls how many bins or data points are displayed in a graphic chart.

      Note:

      Optional: If you select Chart Bins, the Chart Bins dialog opens. You can enter the number of bins to use and can select whether to use the chart range as displayed or the entire chart range, including extreme values excluded from display.

    • Sensitivity Data — Sensitivity data (such as the rank correlation coefficient) for all pairs of assumptions and forecasts indicating the strength of the relationship. Optional: If you plan to extract sensitivity data, select Store Assumption Values For Sensitivity Analysis on the Options tab of the Run Preferences dialog before you run a simulation (Setting Options Preferences). Note: Data is extracted for all assumptions regardless of assumptions selected for extraction.

    • Trial Values — The generated assumption and forecast values for each simulation trial.

    • Capability Metrics — Process capability metric values, if available. If you have activated the process capability features and have generated capability metrics, you can extract them (Extracting Capability Metrics).

      Data types are extracted in the order they are displayed in the Select Data To Extract list. You can use the up and down arrows to rearrange the data types.

  3. In the Forecasts group, select forecasts for data extraction:

    • All includes selected data and restored results for all forecasts in the current simulation.

    • Choose includes selected data for only selected forecasts. Only forecasts for which data was generated or restored are offered for selection.

    • None does not extract any forecast data.

  4. In the Assumptions group, select assumptions for data extraction (All, Choose, or None, as described for forecasts in step 3):

  5. If you have active OptQuest or Predictor data, make appropriate settings to extract the target data. For details, see the Crystal Ball Decision Optimizer OptQuest User's Guide or the Crystal Ball Predictor User's Guide.

  6. Click the Options tab to specify a location or formatting for the extracted data.

  7. In the Location area of the Options tab:

    • To extract data to a new workbook, select New Workbook.

    • To extract data to a new worksheet in the active workbook, select Current Workbook, and then New Sheet.

    • To extract data to the current sheet, select Current Workbook, and then Current Sheet.

  8. Specify the name of the sheet and the first cell of the range where the extracted data will be stored.

  9. Review settings in the Formatting group to indicate how to format the extracted data:

    • Include Labels adds row and column headers to the data table. Otherwise, just the numeric values are extracted.

    • Include Cell Locations adds the workbook, worksheet, and cell address with the object name in the column header. Otherwise, only the object name is displayed.

      This figure displays the statistics and location of cells to be included.
    • AutoFormat applies the following formats to extracted data:

      • Bold font for column headers

      • Border beside row labels

      • Border beneath column headers

      • Border before first assumption

      • Numeric formatting for values

      • AutoFit width to columns

  10. Optional: Click Defaults at any time to restore the original settings to both tabs of the Extract Data dialog.

  11. When the Data and Options tab settings are complete, click OK.

    Crystal Ball extracts the simulation data to the specified worksheet location. The extracted data is arranged as columns of forecasts and assumptions and rows of data. You can sort, modify, print, or save the data in the same way as any other spreadsheet.

For examples of extracted data, see Data Extraction Examples.

Sorting Extracted Data

You can sort extracted data in several ways: by name, by cell row order, or by cell column order. For instructions, see Selecting Assumptions, Forecasts, and other Data Types.