GRASS GIS Temporal Framework

Introduction

The GRASS GIS Temporal Framework implements the temporal GIS functionality of GRASS GIS and provides an API to implement spatio-temporal processing modules. The framework introduces space time datasets that represent time series of raster, 3D raster or vector maps. This framework provides the following functionalities:

  • Assign time stamp to maps and register maps in the temporal database
  • Modification of time stamps
  • Creation, renaming and deletion of space time datasets
  • Registration and un-registration of maps in space time datasets
  • Query of maps that are registered in space time datasets using SQL where statements
  • Analysis of the spatio-temporal topology of space time datasets
  • Sampling of space time datasets
  • Computation of temporal and spatial relationships between registered maps
  • Higher level functions that are shared between modules

Most of the functions described above are member functions of dedicated map layer and space time dataset classes.

Temporal API

The temporal framework API consists of several dedicated modules. Each module contains one or several classes as well as function definition. The API can be roughly divided in a low level and high level part. However, several functions from the low level module are used in module implementation as well.

Low level API

The low level API implements the core functionality of the temporal framework. Core functionality is for example the database interface, the temporal database creation and initialization, the SQL object serialization, all classes that represent table entries, datetime mathematics and many more.

core

The core functionality of the temporal framework:

  • Initialization function init()
  • Definition of global variables
  • Several global functions to access TGIS specific variables
  • Interfaces to the TGIS C-library and PyGRASS messenger objects
  • Database interface connection class SQLDatabaseInterfaceConnection to sqlite3 and postgresql database backends
  • Functions to create the temporal database

base

Implements of basic dataset information and SQL conversion of such information:

  • Definition of the SQL serialize class DictSQLSerializer that converts the content of temporal classes into SQL SELECT, INSERT or UPDATE statements
  • Definition of SQLDatabaseInterface that is the base class for all temporal datatype subclasses
  • Contains classes for all datasets [1] that contain basic information (id, name, mapset, creator, …)

spatial_extent

Implements of 2d and 3d spatial extents of all datasets:

  • Implements class SpatialExtent that is the base class for all dataset specific spatial extent classes It provides spatial topological logic and operations for 2D and 3D extents
  • Implements spatial extent classes for all datasets [1]

temporal_extent

Implements of the temporal extent of all datasets for relative and absolute time:

  • Implements class TemporalExtent that is the base class for all dataset specific temporal extent classes It provides temporal topological logic and operations
  • Implements temporal extent classes for relative time and absolute time for all datasets [1]

metadata

Implements the metadata base classes and datatype specific derivatives fpr all datasets [1].

spatial_topology_dataset_connector

Implements the interface to link datasets by spatial topological relations

temporal_topology_dataset_connector

Implements the interface to link datasets by temporal topological relations

c_libraries_interface

The RPC C-library interface for exit safe and fast access to raster, vector and 3D raster information.

temporal_granularity

The computation of the temporal granularity for a list of AbstractDataset objects for absolute and relative is implemented here.

datetime_math

This module contains function to parse, convert and process datetime objects in the temporal framework.

Spatio-temporal algebra classes for space time raster and vector datasets are defined in:

High level API

The high level API utilizes the low level API. Its classes and functions are usually used to implement temporal processing algorithms and temporal GRASS modules.

abstract_dataset

  • Implements the base class for all datasets [1] AbstractDataset.
  • Implements the the select, insert and update functionality as well as convenient functions to access the base, extent and metadata information

abstract_map_dataset

  • Implements the base class AbstractMapDataset for all map layer specific classes
  • Provides the interface to all map layer specific information in the temporal database

abstract_space_time_dataset

  • Implements the base class AbstractSpaceTimeDataset for all Space Time Datasets classes
  • Contains the creation and deletion functionality, the map registration and un-registration, access methods to map layer objects and so on
  • Provides the interface to all Space Time Dataset specific information in the temporal database

space_time_datasets

This module contains all classes that represent specific datasets [1]. A module developer uses these map layer and Space Time Dataset object representations to perform spatio-temporal tasks.

spatio_temporal_relationships

The logic to compute spatio-temporal topology for a single list or two lists of AbstractDataset objects is implemented in this module. The class SpatioTemporalTopologyBuilder provides a convenient interface for topology computation.

gui_support

Helper functions to support the listing of space time datasets in the automatically generated GUI.

Shared Module functionality

Functionality that is shared between different temporal GRASS modules, such as map listing, space time dataset creation, map registration and un-registration, aggregation, extraction, map calculation, statistics as well as import and export of space time datasets.

aggregation

Aggregation of Space Time Raster Datasets based on topological relations. Used in t.rast.aggregate and t.rast.aggregate.ds

extract

Extraction of subsets from Space Time Datasets including map algebra and vector selection statements. Used in t.rast.extract, t.rast3d.extract and t.vect.extract.

factory

Factory functions to create datasets of all types [1].

open_stds

Convenient functions to open existing Space Time Datasets or to create new ones. Used in almost all temporal modules.

list_stds

Convenient functions to list datasets of all types [1] registered in the temporal database.

mapcalc

Simple temporal algebra for Space Time Raster and 3d Raster datasets. Used in t.rast.mapcalc and t.rast3d.mapcalc

register

Convenient functions to register a single or multiple map layer in the temporal database and SpaceTime Datasets. Used in several modules, most important t.register.

sampling

Sampling functions used in several modules.

stds_export

Functions to export of Space Time Datasets, used in t.rast.export, t.rast3d.export and t.vect.export.

stds_import

Functions to import Space Time Datasets, used in t.rast.import, t.rast3d.import and t.vect.import.

univar_statistics

Simple statistical analysis functions for Space Time Datasets, used in t.rast.univar, t.rast3d.univar and t.vect.univar.
[1](1, 2, 3, 4, 5, 6, 7, 8) : Raster Map Layer, 3d Raster Map Layer, Vector Map Layer, Space time Raster Datasets (STRDS), Space Time 3d Raster Datasets (STR3DS) and Space Time Vector Datasets (STVDS)

Here the full list of all temporal modules:

Examples

Howto start example

This simple example shows how to open a space time raster dataset to access its registered maps.

# Lets import the temporal framework and
# the script framework
import grass.temporal as tgis
import grass.script as grass

# Make sure the temporal database exists
# and set the temporal GIS environment
tgis.init()

# We create the temporal database interface for fast processing
dbif = tgis.SQLDatabaseInterfaceConnection()
dbif.connect()

# The id of a space time raster dataset is build from its name and its mapset
id = "test@PERMANENT"

# We create a space time raster dataset object
strds = tgis.SpaceTimeRasterDataset(id)

# Check if the space time raster dataset is in the temporal database
if strds.is_in_db(dbif=dbif) == False:
    dbif.close()
    grass.fatal(_("Space time %s dataset <%s> not found") % (
        strds.get_new_map_instance(None).get_type(), id))

# Fill the object with the content from the temporal database
strds.select(dbif=dbif)

# Print information about the space time raster dataset to stdout
strds.print_info()

# Get all maps that are registered in the strds and print
# information about the maps to stdout
maps = strds.get_registered_maps_as_objects(dbif=dbif)

# We iterate over the temporal sorted map list
for map in maps:
    # We fill the map object with the content
    # from the temporal database. We use the existing
    # database connection, otherwise a new connection
    # will be established for each map object
    # which slows the processing down
    map.select(dbif=dbif)
    map.print_info()

# Close the database connection
dbif.close()

Creation of a space time dataset

This example shows howto create a space time dataset. The code is generic and works for different space time datasets (raster, 3D raster and vector):

# Lets import the temporal framework and
# the script framework
import grass.temporal as tgis
import grass.script as grass

# The id of the new space time dataset
id="test@PERMANENT"
# The title of the new space time dataset
title="This is a test dataset"
# The description of the space time dataset
description="The description"
# The type of the space time dataset (strds, str3ds or stvds)
type="strds"
# The temporal type of the space time dataset (absolute or relative)
temporal_type="absolute"

# Make sure the temporal database exists
# and set the temporal GIS environment
tgis.init()

# We use the dataset factory to create an new space time dataset instance of a specific type
stds = tgis.dataset_factory(type, id)

# We need a dtabase connection to insert the content of the space time dataset
dbif = tgis.SQLDatabaseInterfaceConnection()
dbif.connect()

# First we check if the dataset is already in the database
if stds.is_in_db(dbif=dbif) and overwrite == False:
    dbif.close()
    grass.fatal(_("Space time %s dataset <%s> is already in the database. "
                    "Use the overwrite flag.") %
                (stds.get_new_map_instance(None).get_type(), name))

# We delete the exiting dataset and create a new one in case we are allowed to overwrite it
if stds.is_in_db(dbif=dbif) and overwrite == True:
    grass.warning(_("Overwrite space time %s dataset <%s> "
                    "and unregister all maps.") %
                (stds.get_new_map_instance(None).get_type(), name))
    stds.delete(dbif=dbif)
    stds = stds.get_new_instance(id)

# We set the initial values. This function also created the command history.
stds.set_initial_values(temporal_type=temporaltype, semantic_type="mean",
                        title=title, description=description)

# Now we can insert the new space time dataset in the database
stds.insert(dbif=dbif)

# Close the database connection
dbif.close()

Temporal shifting

import grass.script as grass
import grass.temporal as tgis

id="test@PERMANENT"
type="strds"

# Make sure the temporal database exists
tgis.init()

dbif = tgis.SQLDatabaseInterfaceConnection()
dbif.connect()

stds = tgis.dataset_factory(type, id)

if stds.is_in_db(dbif) == False:
    dbif.close()
    grass.fatal(_("Space time dataset <%s> not found in temporal database") % (id))

stds.select(dbif=dbif)

stds.snap(dbif=dbif)

stds.update_command_string(dbif=dbif)
dbif.close()

References

Authors:Soeren Gebbert
TODO:add more documentation