Dimensional Moving Statistics function

Available with Image Analyst license.

Available with Spatial Analyst license.

Overview

Calculates statistics over a moving window on multidimensional data along a specified dimension.

Notes

Moving statistics can also be called moving window statistics, rolling statistics, or running statistics. A predefined window around each dimension value is used to calculate various statistics and then move to the next. The Backward Window and Forward Window parameters allow you to define window sizes on both sides of the dimension.

The input raster can only be a multidimensional raster in Cloud Raster Format (.crf file).

Only one dimension will be processed by this function. By default, the first dimension other than x,y will be used as the processing dimension.

The Circular Mean statistics type is calculates the mean for angles or other cyclic quantities, such as compass direction in degrees. When this option is selected, the Circular Wrap Value parameter becomes available. Use this parameter to designate a value to wrap around to calculate the circular mean. For example, for angle calculation, the circular wrap value should be 360 degrees, which means the value 360 will be wrapped to 0, and the value 370 will be wrapped to 10. Another use is for time calculations of months in a year, in which case the circular wrap value should be 12. In this case, an input value of 13 will be wrapped to 1.

If the input multidimensional raster is integer, the valid options for Statistics type are Mean, Circular Mean, Majority, Maximum, Median, Minimum, and Percentile. If the input raster is float, Majority is not available.

If the input raster is integer, the output raster will be integer for the Majority, Maximum, and Minimum statistics. The output will be float for Mean, Circular Mean, Median, and Percentile.

If the input raster is of float type, the output will be float for all of the available statistic types.

For the Majority setting of the Statistics type parameter, if one cell value occurs most frequently in the neighborhood (a single value with the highest frequency of occurrence), that value will be returned for the processing cell. If two or more cell values have the highest frequency of occurrence (a tie among cell values), no single majority value is identified, and the output for that processing cell will be NoData.

Parameters

Parameter nameDescription

Input Raster

(Required)

The input raster can only be a multidimensional raster in Cloud Raster Format (.crf file).

Dimension

The name of the dimension along which the window will move.

The default value is the first dimension other than x,y found in the input multidimensional raster.

Backward Window

The value of how many slices before or above to be included in the defined window. The value must be a positive integer from 1 to 100. The default value is 1.

The unit of this parameter is slice.

Forward Window

The value of how many slices after or below to be included in the defined window. The value must be a positive integer from 1 to 100. The default value is 1.

The unit of this parameter is slice.

NoData Handling

Specifies how NoData values will be handled by the statistic calculation.

  • Data—NoData values in the value input will be ignored in the results of the defined window that they fall within. This is the default.
  • NoData—Output values will be NoData if any NoData values exist in the input within the defined window.
  • Fill NoData—NoData cell values will be replaced using the selected statistic within the defined window.

Statistics Type

Specifies the statistic type to be calculated.

  • Mean—The mean (average value) of the cells in the defined window will be calculated. This is the default.
  • Circular Mean—The mean for angles or other cyclic quantities—such as compass direction in degrees, daytimes, or fractional parts of real numbers—will be calculated. When this statistics type is selected, the Circular Wrap Value parameter becomes available. Use this parameter to designate a wrap value.
  • Majority—The majority (value that occurs most often) of the cells in the defined window will be identified.
  • Maximum—The maximum (largest value) of the cells in the defined window will be identified.
  • Median—The median of the cells in the neighborhood will be identified.
  • Minimum—The minimum (smallest value) of the cells in the neighborhood will be identified.
  • Percentile—A percentile of the cells in the neighborhood will be calculated. When this statistics type is selected, the Percentile Value and Percentile Interpolation Type parameters become available. Use these new parameters to designate the percentile to calculate and choose the interpolation type to use, respectively.

Percentile Value

The percentile that will be calculated when Percentile is selected as the statistics type. The default is 90, for the 90th percentile.

The values can range from 0 through 100. The 0th percentile is essentially equivalent to the minimum statistic, and the 100th percentile is equivalent to the maximum statistic, with the exception that the result will be floating point. A value of 50 will produce the same result as the median statistic.

Percentile Interpolation Type

Specifies the method of interpolation that will be used when the percentile value falls between two cell values.

  • Auto-detect—If the input raster is of integer pixel type, the Nearest method will be used. If the input raster is of float pixel type, the Linear method will be used. This is the default.
  • Nearest—The nearest available value to the percentile will be used. In this case, the output pixel type will be the same as that of the input value raster.
  • Linear—The weighted average of the two surrounding values from the percentile will be used. In this case, the output pixel type will be floating point.

Circular Wrap Value

The value that will be used to convert a linear value to the range of a given circular mean.

Its value must be positive. The default value is 360 degrees.

Related topics


In this topic
  1. Overview
  2. Notes
  3. Parameters