Tileset Creation
Components to create tilesets out of vector and spatial index data.
These components require the CARTO Analytics Toolbox installed in the chosen connection to build the workflow.
Valid table names for tileset's destination:
Provided FQNs need to follow convention on each data warehouse. Invalid names (according to each data warehouse constraints) will make the component fail. Check the documentation for each.
Create H3 Aggregation Tileset
Description
This component creates an aggregation tileset from an input containing H3 cell indexes. On each level of the tileset, cells will be aggregated into their parents, along with their associated properties.
Inputs
Input table: This component expects a table that has at least one column containing H3 indexes.
Settings
Tileset destination: select a destination that is available on your connection to store the resulting tileset table.
H3 column: The column in he input table that contains the H3 index.
H3 input data resolution: It is required to manually enter the actual resolution of the indexes contained in your input.
Minimum H3 tileset resolution: Your tileset will be generated at a range of different H3 resolutions. This setting specifies the minimum resolution in that range. Take into account that H3 resolutions don't match exactly the zoom levels in a web map.
Maximum H3 tileset resolution: Your tileset will be generated at a range of different H3 resolutions. This setting specifies the maximum resolution in that range. Take into account that H3 resolutions don't match exactly the zoom levels in a web map.
Aggregation resolution: On each resolution level in the range set by the minimum and maximum tileset resolution, cells will be aggregated into their parent cells, along with their properties. This setting controls the granularity of that aggregation. For example, let's take a dataset that contains H3 cells of resolution 14. When generating the lower resolution levels of the tileset (bigger cells), the original cells are smaller than a pixel on your screen and can't be visualized. In this case, assuming that 'Aggregation resolution' is set to 5:
at resolution 4, the original cells are going to be aggregated into their level 9 parents (4+5);
at resolution 5, original cells will be aggregated into their level 10 parents (5+5)s;
at resolution 6, original cells will be aggregated into their level 11 parents (6+5).
In general, cells will be aggregated into a resolution level that is the result of: current resolution level + aggregation resolution.
Taking the above into consideration, bear this formula in mind:
Maximum H3 tileset resolution + Aggregation resolution <= H3 input data resolution
If your settings don't follow this rule, you will get an error when generating the tileset.
Properties: Select a property from the input columns and an aggregation method to add aggregated properties to the tileset. The aggregation will be computed for every cell on each resolution level, considering the child cells in the input's data.
Add custom expression: Check this option to define additional custom properties. The custom properties are specified on a JSON object that needs to follow this format:
{ "new_prop_name": { "formula": "COUNT(*)", "type": "Number" }, "most_common_category": { "formula": "APPROX_TOP_COUNT(category, 1)[OFFSET(0)].value", "type": "String" }, ... }The expression informulaneeds to follow your data warehouse's SQL syntax.
Advanced options
Metadata: Use this setting (as a JSON input) to define "name", "description" and "legend" metata objects to be included in the TileJSON. Other custom objects with arbitrary names can also be added, resulting in a "extra_metadata" object in the TileJSON.
Output
This component doesn't have an output that can be connected to other nodes downstream. The generated tileset will be stored on the specified destination table.
Create Point Aggregation Tileset
Description
This component creates a tileset from an input containing Points. It will aggregate points close to each other and their properties.
Inputs
Input table: This component expects a table that has at least one geospatial column containing points of type
GEOMETRYorGEOGRAPHYcolumn depending on the connection provider.
Settings
Tileset destination: select a destination that is available on your connection to store the resulting tileset table.
Geo column: A column in the input table that contains the Points.
Minimum zoom level: Your tileset will be generated at a range of different zoom levels. This setting specifies the lower limit in that range.
Maximum zoom level: Your tileset will be generated at a range of different zoom levels. This setting specifies the higher limit in that range.
Aggregation resolution: On each zoom level in the range set with the options above, points will be aggregated into a grid, along with their properties. This setting controls the granularity of that aggregation. In other words, it controls the the size of the grid cells used to compute the aggregation. For example, assuming the Aggregation resolution is set to 8:
at zoom level 6, the points will be aggregated into a grid where cells are equivalent to tiles of zoom level 14 (6+8).
At zoom level 7, the points will be aggregated into a grid where cells are equivalent to tiles of zoom level 15 (7+8).
At zoom level 8, the points will be aggregated into a grid where cells are equivalent to tiles of zoom level 16 (8+8).
In general, the size of the grid's cells is going to be determined by current zoom level + aggregation resolution.
Aggregation placement: Since points are aggregated into a grid, there are different options for the resulting placement:
cell-centroid: The resulting feature in the tileset is a point a the cell's centroid.
cell: The resulting feature in the tileset is the cell's boundary.
features-any: The resulting feature in the tileset is a random point from the ones that are being aggregated on that cell. This helps avoiding a gridded effect on the visualization
features-centroid: The resulting feature in the tileset is a point that represents the centroid of the points aggregated on that cell. This helps avoiding a gridded effect, while keeping a sense of the distribution of points within the cell.
Properties: Select a property from the input columns and an aggregation method to add aggregated properties to the tileset. The aggregation will be computed for every cell on each resolution level, considering the child cells in the input's data.
Add custom expression: Check this option to define additional custom properties. The custom properties are specified on a JSON object that needs to follow this format:
{ "new_prop_name": { "formula": "COUNT(*)", "type": "Number" }, "most_common_category": { "formula": "APPROX_TOP_COUNT(category, 1)[OFFSET(0)].value", "type": "String" }, ... }The expression informulaneeds to follow your data warehouse's SQL syntax.
Advanced options
Minimum zoom column: A column that controls the minimum zoom level where that specific point should be considered to be included in the tile.
Maximum zoom column: A column that controls the maximum zoom level where that specific point should be considered to be included in the tile.
Maximum features per tile: The maximum number of features that can be included in the tileset. This setting is applied after the aggregation is computed. When the limit is reached, points are going to be dropped from the tile.
This setting is not available in all data warehouses.
Maximum kilobytes per tile: The maximum resulting size in kilobytes. This limit is based on an estimation that only considers the size of the geometries and not the included properties. When the limit is reached, points are going to be dropped from the tile.
This setting is not available in all data warehouses.
Tile feature order: When any of the limits above is reached, points are dropped from the tile. This setting expects a SQL expression to define the order of priority on the resulting points. The higher in the ordering, the less likely to be dropped.
Metadata: Use this setting (as a JSON input) to define "name", "description" and "legend" metata objects to be included in the TileJSON. Other custom objects with arbitrary names can also be added, resulting in a "extra_metadata" object in the TileJSON.
Output
This component doesn't have an output that can be connected to other nodes downstream. The generated tileset will be stored on the specified destination table.
Create Quadbin Aggregation Tileset
Description
This component creates an aggregation tileset from an input containing Quadbin cell indexes. On each level of the tileset, cells will be aggregated into their parents, along with their associated properties.
Inputs
Input table: This component expects a table that has at least one column containing H3 indexes.
Settings
Tileset destination: select a destination that is available on your connection to store the resulting tileset table.
Quadbin column: The column int he input table that contains the Quadbin index.
Quadbin input data resolution: It is required to manually enter the actual resolution of the indexes contained in your input.
Minimum Quadbin tileset resolution: Your tileset will be generated at a range of different Quadbin resolutions. This setting specifies the minimum resolution in that range. Take into account that Quadbin resolutions levels are equivalent to zoom levels on a CARTO map.
Maximum Quadbin tileset resolution: Your tileset will be generated at a range of different Quadbin resolutions. This setting specifies the maximum resolution in that range.
Aggregation resolution: On each resolution level in the range set by the minimum and maximum tileset resolution, cells will be aggregated into their parent cells, along with their properties. This setting controls the granularity of that aggregation. For example, let's take a dataset that contains Quadbin cells of resolution 14. When generating the lower resolution levels of the tileset (bigger cells), the original cells are smaller than a pixel on your screen and can't be visualized. In this case, assuming that 'Aggregation resolution' is set to 5:
at resolution 4, the original cells are going to be aggregated into their level 9 parents (4+5);
at resolution 5, original cells will be aggregated into their level 10 parents (5+5)s;
at resolution 6, original cells will be aggregated into their level 11 parents (6+5).
In general, cells will be aggregated into a resolution level that is the result of: current resolution level + aggregation resolution.
Taking the above into consideration, bear this formula in mind:
Maximum Quadbin tileset resolution + Aggregation resolution <= Quadbin input data resolution
If your settings don't follow this rule, you will get an error when generating the tileset.
Properties: Select a property from the input columns and an aggregation method to add aggregated properties to the tileset. The aggregation will be computed for every cell on each resolution level, considering the child cells in the input's data.
Add custom expression: Check this option to define additional custom properties. The custom properties are specified on a JSON object that needs to follow this format:
{ "new_prop_name": { "formula": "COUNT(*)", "type": "Number" }, "most_common_category": { "formula": "APPROX_TOP_COUNT(category, 1)[OFFSET(0)].value", "type": "String" }, ... }The expression informulaneeds to follow your data warehouse's SQL syntax.
Advanced options
Metadata: Use this setting (as a JSON input) to define "name", "description" and "legend" metata objects to be included in the TileJSON. Other custom objects with arbitrary names can also be added, resulting in a "extra_metadata" object in the TileJSON.
Output
This component doesn't have an output that can be connected to other nodes downstream. The generated tileset will be stored on the specified destination table.
Create Vector Tileset
Description
This component creates a tileset from an input containing Point, Lines or Polygons.
Inputs
Input table: This component expects a table that has at least one geospatial column of type
GEOMETRYorGEOGRAPHYdepending on the connection provider.
Settings
Tileset destination: select a destination that is available on your connection to store the resulting tileset table.
Geo column: A column in the input table that contains the Points.
Minimum zoom level: Your tileset will be generated at a range of different zoom levels. This setting specifies the lower limit in that range.
Maximum zoom level: Your tileset will be generated at a range of different zoom levels. This setting specifies the higher limit in that range.
Properties: Select a list of properties from the input columns.
Advanced options
Minimum zoom column: A column that controls the minimum zoom level where that specific point should be considered to be included in the tile.
Maximum zoom column: A column that controls the maximum zoom level where that specific point should be considered to be included in the tile.
Maximum features per tile: The maximum number of features that can be included in the tileset. This setting is applied after the aggregation is computed. When the limit is reached, points are going to be dropped from the tile.
This setting is not available in all data warehouses.
Maximum kilobytes per tile: The maximum resulting size in kilobytes. This limit is based on an estimation that only considers the size of the geometries and not the included properties. When the limit is reached, points are going to be dropped from the tile.
This setting is not available in all data warehouses.
Tile feature order: When any of the limits above is reached, points are dropped from the tile. This setting expects a SQL expression to define the order of priority on the resulting points. The higher in the ordering, the less likely to be dropped.
Generate this tileset as a custom boundary for CARTO + deck.gl: Check this option if your tileset is going to be used as a boundary. Learn more about Boundaries.
Metadata: Use this setting (as a JSON input) to define "name", "description" and "legend" metata objects to be included in the TileJSON. Other custom objects with arbitrary names can also be added, resulting in a "extra_metadata" object in the TileJSON.
Output
This component doesn't have an output that can be connected to other nodes downstream. The generated tileset will be stored on the specified destination table.
Last updated