HomeAcademyLoginTry for free

Statistics

Components to perform spatial statistics calculations.

These components require the CARTO Analytics Toolbox installed in the chosen connection to build the workflow.

Cluster Time Series

Description

This component computes a clusterization based on the values of a feature over time (i.e. sales of different stores, temperature in different H3 cells, etc) and assigns a cluster number to each, using different methods to evaluate the values and their distribution over time.

Inputs

  • Input table: This component expects a table containing a column with at least a timestamp column, a value column and an index to be used as partition (usually an spatial index or a feature id)

Settings

  • Timestamp column: Select a column in the input that contains timestamps/dates.

  • Partition column: Each unique value on this column will be assigned a cluster.

  • Value column: Clusters will be calculated based on the value of this column along time

  • Number of clusters: Define the number of clusters to be generated.

  • Method:

    • Value

    • Profile

Outputs

  • Output table: This component will generate a table with the following columns:

    • Partitioning column with the same name as the given column from inputs.

    • Cluster (String): Contains the assigned cluster number.

Composite Score Supervised

Description

This component derives a spatial composite score as the residuals of a regression model which is used to detect areas of under- and over-prediction. The response variable should be measurable and correlated with the set of variables defining the score. For each data point, the residual is defined as the observed value minus the predicted value. Rows with a NULL value in any of the input variables are dropped.

Inputs

  • Input table

  • Column with unique geographic identifier (spatial index)

  • Input variables

  • Response variables

  • Model regressor

  • Model options

  • Output formatting

  • R-squared threshold

Outputs

  • Output table

External links

BigQuery reference

Snowflake reference

Composite Score Unsupervised

Description

This component combines (spatial) input variables into a meaningful composite score. The composite score can be derived using different methods, scaling and aggregation functions and weights. Rows with a NULL value in any of the model predictors are dropped.

Inputs

  • Input table

  • Column with unique geographic identifier (spatial index)

  • Input variables

  • Scoring method

    • Entropy

    • Custom Weights:

      • Variables and weights:When defining custom weights, you will find a button that controls the direction of the weight:

        • When specifying custom weights, the component will internally apply a normalization so that the sum of all weights is equal to 1

        • Scaling function:

          • MIN_MAX_SCALER: data is rescaled into the range [0,1] based on minimum and maximum values. Only numerical variables are allowed.

          • STANDARD_SCALER: data is rescaled by subtracting the mean value and dividing the result by the standard deviation. Only numerical variables are allowed.

          • RANKING: data is replaced by its percent rank, that is by values ranging from 0 lowest to 1. Both numerical and ordinal variables are allowed (categorical and boolean variables need to be transformed to ordinal).

          • DISTANCE_TO_TARGET_MIN(_MAX,_AVG):data is rescaled by dividing by the minimum, maximum, or mean of all the values. Only numerical variables are allowed.

          • PROPORTION: data is rescaled by dividing by the sum total of all the values. Only numerical variables are allowed.

        • Aggregation Function:

          • LINEAR: the spatial composite is derived as the weighted sum of the scaled individual variables.

          • GEOMETRIC: the spatial composite is given by the product of the scaled individual variables, each to the power of its weight.

      • First PC

        • Correlation variable: the spatial score will be positively correlated with the selected variable (i.e. the sign the spatial score is set such that the correlation between the selected variable and the first principal component score is positive).

        • Correlation threshold: the minimum absolute value of the correlation between each individual variable and the first principal component score.

  • Output formatting

    • None

    • Equal Intervals

    • Jenks

    • Quantiles

    • Return Range

Outputs

  • Result table

  • Lookup table:only available if output formatting is Equal Intervals, Jenks, or Quantiles

External links

BigQuery reference

Snowflake reference

Cronbach Alpha Coefficient

Description

This component computes the Cronbach Alpha Coefficient, which can be used to measure internal consistency of different variables.

This component can be used to determine wether a combination of variables are consistent enough to be used to create a Composite Score using one of the components above.

Inputs

  • Input table

  • Input variables

Outputs

  • Output table

External links

BigQuery reference

Snowflake reference

Getis Ord

Description

This component computes the Getis-Ord Gi* statistic for each spatial index in the source table.

Inputs

  • Source table [Table]

  • Index column [Column]

Make sure that the input data for this component doesn't contain any NULL values on the index column. Otherwise the execution will fail.

  • Value column [Column]

  • Kernel function for spatial weights [Selection]

  • Size [Number]

Outputs

  • Result table [Table]

External links

Getis Ord Spacetime

Description

This component computes the spacetime Getis-Ord Gi* statistic for each spatial index and datetime timestamp in the source table.

Inputs

  • Source table [Table]

  • Index column [Column]

Make sure that the input data for this component doesn't contain any NULL values on the index column. Otherwise the execution will fail.

  • Date column [Column]

  • Value column [Column]

  • Kernel function for spatial weights [Selection]

  • Kernel function for temporal weights [Selection]

  • Size [Number]

  • Temporal bandwidth [Number]

  • Time interval [Selection]

Outputs

  • Result table [Table]

External links

GWR

Description

This component runs a spatial-index-based Geographically Weighted Regression (GWR) model.

Inputs

  • Input table [Table]

  • Index column [Column]

  • Feature variables [Column]

  • Target variable [Column]

  • K-ring size [Number]

  • Kernel function [Selection]

  • Fit intercept [Boolean]

Outputs

  • Result table [Table]

External links

Hotspot Analysis

This component requires the CARTO Analytics Toolbox installed in the chosen connection to build the workflow.

Description

This component is used to locate hotspot areas by calculating a combined Getis-Ord Gi* statistic using a uniform kernel over one or several variables.

Inputs

  • Source table [Table]

  • Index column [Column]

  • Input variables [Column]

  • Variable weights [String]

  • K-ring size [Number]

  • Significance level [Number]

Outputs

  • Result table [Table]

External links

Local Moran's I

Description

This component computes the local Moran's I statistic for each spatial index in the source table.

Inputs

  • Source table [Table]

  • Index column [Column]

  • Value column [Column]

  • Size [Number]

  • Decay function [Selection]

  • Permutations [Number]

Outputs

  • Result table [Table]

External links

Moran's I

Description

This component computes the Moran's I statistic for each spatial index in the source table.

Inputs

  • Source table [Table]

  • Index column [Column]

  • Value column [Column]

  • Size [Number]

  • Decay function [Selection]

Outputs

  • Result table [Table]

External links

Spacetime Hotspots Classification

Description

This component takes the output of Getis Ord Spacetime and classifies each location into specific types of hotspots or coldspots, based on patterns of spatial clustering and intensity trends over time.

Inputs

  • Input table: This component expects to be connected to the output of a Getis Ord Spacetime successfully executed node, or a table that contains that same result.

Settings

  • Index column: Select a column in the input that contains a spatial index (Quadbin or H3).

  • Date column: Select a column in the input that contains date/timestamp.

  • Gi Value: Select a column in the input that contains a Gi value generated by a Getis Ord Spacetime component.

  • P Value: Select a column in the input that contains a P value generated by a Getis Ord Spacetime component.

  • Threshold: Select the threshold of the P value for a location to be considered as hotspot/coldspot.

  • Algorithm: Select the algorithm to be used for the monotonic trend test:

    • Mann-Kendall (default)

    • Modified Mann-Kendall

Output

  • Output table: The component will return a table with the column that contains the index and the resulting columns from the procedure's execution.

PreviousSpatial OperationsNextTileset Creation

Last updated