Generalizer

There are two typical reasons for using Process When Group Changes (Advanced) . The first is incoming data that is intended to be processed in groups (and is already so ordered). In this case, the structure dictates Group By usage - not performance considerations.

The second possible reason is potential performance gains.

Performance gains are most likely when the data is already sorted (or read using a SQL ORDER BY statement) since less work is required of FME. If the data needs ordering, it can be sorted in the workspace (though the added processing overhead may negate any gains).

Sorting becomes more difficult according to the number of data streams. Multiple streams of data could be almost impossible to sort into the correct order, since all features matching a Group By value need to arrive before any features (of any feature type or dataset) belonging to the next group. In this case, using Group By with Process At End (Blocking) may be the equivalent and simpler approach.

Note: Multiple feature types and features from multiple datasets will not generally naturally occur in the correct order.

As with many scenarios, testing different approaches in your workspace with your data is the only definitive way to identify performance gains.

Douglas

The Douglas-Peucker algorithm will remove vertices which cause a deviation of less than the Generalization Tolerance, but the location of remaining vertices are not altered. Thus, this algorithm is good at reducing the number of points in a line, it is not very good at preserving the shape or the spatial relationship of the line relative to other entities. When used on polygons, the start point is never removed.

Corresponding parameters:

• Generalization Tolerance
• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments

Douglas With Arc Fitting

The Douglas With Arc Fitting algorithm performs the same operations as the Douglas algorithm, but it will also attempt to fit circular arcs to the data, often resulting in fewer vertices. Any arcs added will be within the Generalization Tolerance of the original input. This algorithm will always remove a number of points that is more than or equal to the number that would be removed by Douglas, however, the result may be an IFMEPath instead of an IFMELine.

Corresponding parameters:

• Generalization Tolerance
• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments

Thin

The Thin algorithm will remove vertices that are less than the Generalization Tolerance distance away from an adjacent vertex. The begin and end points are never moved, unless the entire length of the feature being thinned is less than the tolerance, in which case the feature is replaced by a point feature holding the final coordinate.

Corresponding parameters:

• Generalization Tolerance

Thin No Point

The Thin No Point algorithm will remove vertices that are less than the Generalization Tolerance distance away from an adjacent vertex. The begin and end points are never moved, even when the entire length of the feature being thinned is less than the tolerance, in which case the feature is replaced by a linear feature connecting the first point to the last point.

Corresponding parameters:

• Generalization Tolerance
• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments

Deveau

The Deveau algorithm removes vertices which contribute less to the overall shape of the feature, and may introduce new vertices at positions not originally in the feature as it works. The inherent behavior of the algorithm is such that it invalidates the z coordinate of the vertices, and any measures. Therefore the output features will always be 2D, and have no measures on them. It requires the Smoothness Factor parameter and the Sharpness Angle parameter to be specified.

Corresponding parameters:

• Generalization Tolerance
• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments
• Smoothness Factor
• Sharpness Angle

Wang

The Wang algorithm will iteratively combine, eliminate and exaggerate bends until the input line feature has no bend that is smaller than the given tolerance value.

Corresponding parameters:

• Generalization Tolerance
• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments

McMaster

The McMaster algorithm calculates a new location for each point by first taking the average value of the x and y coordinates of the point and a number of neighboring points. It then slides the averaged point towards the original point according to a specified displacement value. The overall effect is that each point will be pulled towards its neighboring points.

Corresponding parameters:

• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments
• Number of Neighbors
• Displacement Percentage

McMaster Weighted Distance

The McMaster Weighted Distance algorithm performs the same operations as the McMaster algorithm only it uses inverse distance weighting to take into account the distance from each neighbor to the point being moved. The overall effect is that points further away will have less "pull" than points close by.

The Weighting Power parameter is used by the McMaster Weighted Distance algorithm only. It is used to determine the weight of each neighboring point.

Note: For lines, the McMaster algorithms do not change the first and last N points (where N is the number of neighbors), because they don't have enough neighbors for the averaging calculations to work with. For polygons, a wrap-around is used so each point in a polygon will be changed. In the case of adjacent polygons and the Preserve Shared Boundaries option, collinear portions of their boundaries will be smoothed together. The remaining parts of their boundaries will be smoothed as lines. This means that no wrap-around will be used for adjacent polygons.

Corresponding parameters:

• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments
• Number of Neighbors
• Displacement Percentage
• Weighting Power

NURBfit

The NURBfit algorithm will fit lines using B-Spline curves of given polynomial degree. The resulting lines will follow these curves with given segment length. The higher the degree, the smoother the line. An example of usage is smoothing contour lines to remove spikes and simulate the work of a cartographic craftsman.

Corresponding parameters:

• Preserve Shared Boundaries
• Shared Boundaries Tolerance
• Preserve Path Segments
• Degree of Basis Polynomial
• Segment Length

Fillet

The Fillet algorithm will replace linear corners in a line with arc corners with the given fillet radius. If an arc with the specified fillet radius would not fit on a corner, that corner will be skipped.

Inflection Points

The Inflection algorithm will calculate the location of the inflection points along a line and return the list of these points. Inflection points are measures of the sinuosity of a line.

Corresponding parameters:

• Number of Neighbors

Orthogonal Distance Regression

This algorithm replaces the feature's geometry with a line that minimizes the orthogonal distance between it and the original geometry's points. Orthogonal distance means the shortest (perpendicular) distance between a point and a line.

Corresponding parameters:

• None