In this lesson,

we will address the question of topological spatial queries,

so something that completes the geometric spatial queries
that we saw in the previous lesson. These topological spatial queries
are therefore interested in extracting information which characterize not the
spatial objects
themselves but rather the spatial relations
and the relations between spatial objects. The objective of this lesson
is to explore the syntax field of SQL queries
which allow to extract information characterizing spatial relations
between objects hosted in a database. At the end of this lesson
you should be able to write queries
which will allow you to test spatial relations between objects,
to calculate characteristics such as the distance between objects
or to make new objects by intersection or difference
or other operations involving existing spatial objects. Just as in the case of
geometric spatial queries,
topological spatial queries are numerous. There are many keywords that can be used.
We will present only a very small part of
it so this is the reason why I encourage you
to take a look on the sites that document the available functions
in SpatiaLite and in Postgis. Again, the syntax is pretty much the same
in both cases with, in the case of Postgis,
the addition of the letters S and T underlined
before the function. In this lesson,
we will first go back to the notion of topology and topological query,
notion which had been presented at the beginning of the course. Then we will
present the test functions
before addressing the distance function and ending by spatial operators. We had
seen
by presenting the notions of topologies that a relation, a spatial dependence
is defined by a property shared by two spatial entities. As for example the
simplest one of them,
the distance that separates two points. We also saw
that the topological relations are a sub category
of spatial relation, of a qualitative nature,
independent of the measure and characterized by the fact
of being invariant under continuous deformation. The main forms of topological
relations
are the adjacency, the connectivity, the inclusion and the intersection. The
requests which are on the spatial relations
between objects and more particularly on topological relations
can be subdivided roughly into 3 categories. First, the test requests
which return "true" or "false" values. Distance queries
which return a numeric value and action requests
which return spatial objects. Test requests
or test function allow to test spatial relations
or more precisely the topological relations between two spatial entities
on the basis of their geometry. The syntax is always
the SELECT keyword, the test function
with the two geometries concerned in brackets,
separated by a comma, the FROM clause
and the table. The main functions that can be used
are listed here in particular, we can test the equality between
two geometries, the fact that they are distant from one another,
that they touch each other, that one is contained into the other
or that it contains the other, that they overlap,
that they cross each other, etc. In this example
we will select the hotels the name of the hotels
which are located within the Takamaka districts. To do this,
the syntax is SELECT, the name of the hotel, the FROM clause with the two
hotels and districts tables and a conditional clause
in which we specify the fact that the geometry of the hotels
must be inside the geometry of the districts
and that the district name should resemble Takamaka. By executing this query
we get the 6 hotels which are in this district. We can now, in a second example,
replace the hotels by the roads
and search for the roads which intersect the district of Takamaka. So we see that
we have a series of 21 roads
which form an intersection with that district. To store the result of this query
as a new layer we have to add in the request
a "select the geometry attribute" column and define with the geometry name,
which is the field that we will use to make the layer in QGIS,
give a name to this layer, in this case INTERSECT,
and execute this request. We make a second request
to select this time the roads that touch the district of Takamaka. We can see that
this request
does not give any result, so there is no object
which properly touches the district and a third request
to identify the roads which are contained in the district. We also export the
result of that query
as a layer in QGIS and if we now look at these two layers,
the "intersection" layer in red and the "content" layer in yellow,
it is clear that the intersection layer includes more objects
than the layer of what is strictly contained in the district. The distance function
simply allows to calculate the minimum distance separating two geometries,
generally expressed in meters, well it depends on the projection system used. The
syntax is of the same nature as in the
previous cases, the SELECT keyword, the distance function,
the two geometries concerned in brackets, separated by a comma,
the FROM clause, and the name of the table. In this example,
we will build a query to classify the hotels
according to their distance from the road network. In the syntax, we will start
with the SELECT
keyword, the name of the hotels
and a second factor which will be the distance
between the hotel geometries and the geometry of the roads
and we will take the minimum value of all the distances separating a hotel from
all the roads. The field "calculate the minimum distance"
is given an alias, DIST,
in the FROM clause the two tables concerned,
so hotels and roads , 183 00:07:49,807 --> 00:07:52,526 a grouping clause,
the name of the hotels and then a ranking
in increasing order of distances. By executing this query
we find the series of 124 hotels with the distance which separates them
from the road network, between 10 and 600m. A series of spatial operators
to finish, so operators that return new geometric objects
created from existing geometric objects. The first of these operators
is the union query which allows to merge two geometries
into a single object. The intersection request
which returns the common part of two geometries. The difference request
which subtracts a geometry of another. The syntax, as usual
is composed by the SELECT keyword Followed by the function,
the two geometries concerned in brackets, separated by a comma,
the AS GEOMETRY alias to be able to reuse this result,
say that it is a geometry and the table from where the data come. As in the case of
the length of the polylines,
the union request, the UNION keyword,
is part of a series of keywords reserved in SpatiaLite,
that is why we use GUnion instead
and as always in Postgis the equivalent will be ST underlined UNION. Two other
important space operators,
SIMPLIFY and BUFFER. The first, SIMPLIFY,
refers to a simplified geometry, taking into account a tolerance
which is defined by the numerical argument given in the function,
so it is a simplification which is based on a Douglas-Peucker algorithm,
which allows to reduce the number of points of which a geometry is
composed, which can be important
in the display performances of the geometry on the screen. The BUFFER function
returns a geometry which covers an entire area
located at a distance inferior or equal to the number given as an attribute
of an original geometry. It is the concept of buffer zone. At the syntax level
still the SELECT keyword, the function,
the geometry, comma, the numeric variable, a tolerance in the case of SIMPLIFY,
a distance in the case of BUFFER, AS GEOMETRY because we create geometries
and the original table. In this first example
we will illustrate the use of the SIMPLIFY function
by simplifying the road network of the Seychelles with a tolerance of 50. We see
that the request produces
geometric objects indeed. They have to be defined as geometries
to be able to recover them in QGIS as usual. A name is given to the table,
SIMPLIFY, the geometry being of course the GEOMETRY field. This simplified layer
appears here in white and if we go a little bit closer
we see that the detail of the road network has been erased
in favor of a simpler structure. Second example
with the union. We will establish a union request
between two districts, the district of Port Glaud
and the district of Grand'Anse Mahé. We transfer the union request
on the geometry field of the district table and in the conditional clause
which clarifies things we say that we want the name
of districts chosen for the union to be either Port Glaud or Grand'Anse Mahé. The
execution of this request
causes the merging of the two entities. Third case,
we will build a buffer zone, a buffer,
around the Sans Soucis road which is located in the center-western part
of the Seychelles. So we define a buffer with a width of 100m
around that road and in the conditional clause
we will simply say that we want all the segments of the road
which look like Sans Soucis road. The execution of this request
provides geometric objects. With the AS GEOMETRY
and by defining a table name, we add this request
as a layer in the QGIS interface
and we see that at the time I execute the request
the buffer zone is displayed on the map, here in pink,
and if we look a little closer we see that the different sections of the
road will give rise to a succession
of buffer zone pieces which fit with each other. As a final example,
we will create a buffer zone around the two districts
of Port Glaud and Grand'Anse Mahé by taking up the merging request
of these elements from before and this time we will add
the creation of a buffer to this merging. If we move the district layer
above this buffer and that we look little closer
we see that we have created a 100m block
around these two merged districts. In this lesson,
after talking a little bit about the notions of topology
we have seen a few examples of test functions
which allow to evaluate the existence of topological spatial relations
between spatial entities. We have seen an example of using
the distance function To sort hotels according to their distance
to the road network and we have seen some examples
of application of spatial operators which allow to create new objects
by intersection by difference, by creation of buffer zones
and other.