Missing Values

In complex datasets, observations may be missing at some locations and/or some times. This may occur when data for a few locations were lost or deemed unreliable.  This can also happen when some variables are not observed at all time intervals covered by other datasets in the same geography.  Text files and Excel files with missing values can show these as empty placeholders or cells. These can also be identified using a missing value code.  If you are working with DBF files (as part of a shapefile, or as a file for importing datasets), you MUST set a missing value code.  

Click here to see information on representing missing values in the Map View.

Missing values in imported DBF files

The DBF file does not allow empty cells to represent missing data, so it must be defined as a no-data or missing value code. By convention, filling the field with -9's is a common missing value code. To make sure that the data are not analyzed as "-9,999" rather than missing, you need to identify the code so that SpaceStat will handle the missing value code appropriately.  Note that if empty fields in a DBF are not filled, your data may still import, but it will be incorrect, as values may shift to fill the empty space.

Missing values when exporting data

Missing values are shown in SpaceStat tables as a dash. When you export data with missing values, such as Local Moran statistics for geographies that include islands (polygons without neighbors), you need to specify a missing value code for export to a DBF.

Choosing a missing value code

Because the exported DBF does not allow blank entries, you must provide some value to represent the data that are missing. The missing value code should be a value that could not possibly show up as a true data value in the data set. Often, codes such as "-9999" are used so that the code is easy to recognize when you scan a column of data. For example, it is recommended to use negative values if the dataset contains only positive values. Any integer value can be used, including negative numbers. Currently, decimal values and text strings (such as "no data") cannot be used.

Missing values over time

When you import several time slices to become one space-time dataset in SpaceStat, you can import a file for each time slice. Columns with the same name in each data file are considered to be the same variable. Variables that do not appear in all these files are considered to be missing values when excluded.