Missing Data

There are a number of scenarios where you might want to use missing data. When observations are recorded at some locations but not others for a variable, you may want to designate the locations which weren’t observed as missing for that variable. If location data is recorded for some time intervals but not others for a variable, you can designate the values for those time intervals where there is no data as missing. Also if there was an error in data collection or recording you may designate a value as missing. If data is not released for confidentiality reasons, it can be assigned as missing data.

Data can be missing at a location for none, one, all, or any subset of variables. Each variable is treated independently.

Read our blog post for more background information about Missing Data.

How does Vesta handle missing data?

Vesta handles missing data differently in visualizations and calculations. In the maps, Vesta will display an object that is missing for a variable with a custom appearance, often gray by default. In the Table Visualization, Vesta will describe the variable value as missing or abbreviate it with ‘-’.

In calculations, charts, and statistics any missing observations will be excluded from the analysis. These locations will not be included in sample sizes and totals used in calculations. Analyses, results, and descriptive statistics will appear as if these locations did not exist. Note that this applies only to the time interval(s) when the data is missing and the locations will be used at other time intervals when they are not missing.

Prior to importing a data file with missing values, the "empty cells" can be left empty or be assigned a value (often -9999) so that Vesta can identify these as missing values.

Import data file with missing data

The easiest way to specify missing data is upon importing a data file. On the main dialog to import a file, check the box under Data Options to select that the Dataset contains missing data. You can then choose whether the missing data is represented by no data or a specific value. No data includes cases such as when a spreadsheet cell is empty for a missing value. Sometimes files encode missing data with a specific value, for example -9999. If you choose to specify missing data upon import, it must be represented consistently across all variables within that file.

In future versions of Vesta you will also be able to specify missing data after the file has been imported for cases when you have a more complex representation or if you forgot to specify it during the import step.