Difference between revisions of "Advanced Input"

Revision as of 15:05, 12 October 2018

This page explains how to use advanced input definitions when developing an algorithm. These indications are valid for algorithms developed using both the Statistical Algorithms Importer (SAI) and DataMiner.

Spatial Data

Defining Spatial Data

Spatial Data input are passed to an algorithm using Well-Known Text format (WKT), i.e. a text markup language for representing vector geometry objects on a map, spatial reference systems of spatial objects and transformations between spatial reference systems.^[1]

When an algorithm input is defined, a WKT input can be indicated by using proper annotations in the input descriptions. In particular, the annotations highlighted in red in the following table correspond to different geometries. For example a description for an input point could be "Location of an occurrence record [WKT_POINT]".

The currently produced CRS is EPSG:4326 (Examples are still in EPSG:3857).

Spatial Data on the DataMiner Interface

By using a specific widget, DataMiner is able to build a selection interface suited for the WKT input. This process is done by reading the annotation in the input description. The widget allows to easily select the area of interest and reports the WKT selected in a text area. In the end, the WKT string is passed to the algorithm:

Spatial Data, DataMiner

WKT Point, DataMiner

WKT LineString, DataMiner

WKT Polygon, DataMiner

WKT Triangle, DataMiner

WKT Square, DataMiner

WKT Pentagon, DataMiner

WKT Hexagon, DataMiner

WKT Box, DataMiner

WKT Circle, DataMiner

Temporal Data

Temporal Data

Temporal Data are automatically interpreted using specific tags in the description of the inputs. The tags highlighted in red in the following table should be used to indicate the kind of temporal information to associate to the input. Currently, dates and time stamps are supported. For example a description for an input date could be "Date of the occurrence record [DATE]".

Temporal Data on DataMiner

A widget on DataMiner is able to transform the annotations into a selection panel. In the following, two examples are reported:

Temporal Data, DataMiner

Long strings and TextArea

TextArea in SAI

DataMiner for string parameters uses a TextField widget as default. There are cases where it is more useful to use a TextArea widget. To enable TextArea, just add the [TEXTAREA] tag to the description of the parameter in SAI.

TextArea in DataMiner

In the following, what is shown in DataMiner:

Long string and TextArea, DataMiner

NetCDF

NetCDF in SAI

NetCDF (.nc) file information view is supported. There are cases where it is useful to see the information contained in the NetCDF files. To enable the NetCDF Preview, just add the [NETCDF] tag to the description of the parameter in SAI.

NetCDF in DataMiner

In the following, what is shown in DataMiner:

NetCDF file selection, DataMiner

After NetCDF file is selected:

NetCDF file selected, DataMiner

After info icon button is pressed:

NetCDF Preview, DataMiner

Other Inputs

Supported by SAI:

• File Input: [a http link to a file] (Example: FEED_FORWARD_A_N_N_DISTRIBUTION)

• List of strings: [a sequence of values separated by | , Enumerated for SAI ] (Example: OCCURRENCE_ENRICHMENT)

Supported only by Java APIs(no SAI) see How-to Implement Algorithms for DataMiner:

• Table Input: [a http link to a table ] (Example: DBSCAN)

• List of Tables: [a sequence of http links separated by | , each indicating a table ] (Example BIOCLIMATE_HCAF)

• List of columns from a table: [a sequence of names of columns from <NameOfTheTableParameter> separated by | ] (Example: FEED_FORWARD_A_N_N_DISTRIBUTION)

• Column name from a table: [the name of a column from <NameOfTheTableParameter> ] (Example: OCCURRENCE_ENRICHMENT)

References

1. ↑ "Well-known text", Wikipedia