Statistics of FeatureCollection Columns

To reduce properties of features in a FeatureCollection, use featureCollection.reduceColumns(). Consider the following toy example:

Code Editor (JavaScript)

// Make a toy FeatureCollection.
var aFeatureCollection = ee.FeatureCollection([
  ee.Feature(null, {foo: 1, weight: 1}),
  ee.Feature(null, {foo: 2, weight: 2}),
  ee.Feature(null, {foo: 3, weight: 3}),
]);

// Compute a weighted mean and display it.
print(aFeatureCollection.reduceColumns({
  reducer: ee.Reducer.mean(),
  selectors: ['foo'],
  weightSelectors: ['weight']
}));

Python setup

See the Python Environment page for information on the Python API and using geemap for interactive development.

import ee
import geemap.core as geemap

Colab (Python)

# Make a toy FeatureCollection.
a_feature_collection = ee.FeatureCollection([
    ee.Feature(None, {'foo': 1, 'weight': 1}),
    ee.Feature(None, {'foo': 2, 'weight': 2}),
    ee.Feature(None, {'foo': 3, 'weight': 3}),
])

# Compute a weighted mean and display it.
display(
    a_feature_collection.reduceColumns(
        reducer=ee.Reducer.mean(), selectors=['foo'], weightSelectors=['weight']
    )
)

Note that the inputs are weighted according to the specified weight property. The result is therefore:

mean: 2.333333333333333

As a more complex example, consider a FeatureCollection of US census blocks with census data as attributes. The variables of interest are total population and total housing units. You can get their sum(s) by supplying a summing reducer argument to reduceColumns() and printing the result:

Code Editor (JavaScript)

// Load US census data as a FeatureCollection.
var census = ee.FeatureCollection('TIGER/2010/Blocks');

// Filter the collection to include only Benton County, OR.
var benton = census.filter(
  ee.Filter.and(
    ee.Filter.eq('statefp10', '41'),
    ee.Filter.eq('countyfp10', '003')
  )
);

// Display Benton County census blocks.
Map.setCenter(-123.27, 44.57, 13);
Map.addLayer(benton);

// Compute sums of the specified properties.
var properties = ['pop10', 'housing10'];
var sums = benton
    .filter(ee.Filter.notNull(properties))
    .reduceColumns({
      reducer: ee.Reducer.sum().repeat(2),
      selectors: properties
    });

// Print the resultant Dictionary.
print(sums);

Python setup

See the Python Environment page for information on the Python API and using geemap for interactive development.

import ee
import geemap.core as geemap

Colab (Python)

# Load US census data as a FeatureCollection.
census = ee.FeatureCollection('TIGER/2010/Blocks')

# Filter the collection to include only Benton County, OR.
benton = census.filter(
    ee.Filter.And(
        ee.Filter.eq('statefp10', '41'), ee.Filter.eq('countyfp10', '003')
    )
)

# Display Benton County census blocks.
m = geemap.Map()
m.set_center(-123.27, 44.57, 13)
m.add_layer(benton)
display(m)

# Compute sums of the specified properties.
properties = ['pop10', 'housing10']
sums = benton.filter(ee.Filter.notNull(properties)).reduceColumns(
    reducer=ee.Reducer.sum().repeat(2), selectors=properties
)

# Print the resultant Dictionary.
display(sums)

The output is a Dictionary representing the aggregated property according to the specified reducer:

sum: [85579,36245]

Note that the above example uses the notNull() filter to include only features with non-null entries for selected properties in the collection being reduced. It is good practice to check for null entries to catch unexpected missing data and avoid errors resulting from calculations that include null values.

Also note that unlike imageCollection.reduce(), in which reducers are automatically repeated for each band, reducers on a FeatureCollection must be explicitly repeated using repeat(). Specifically, repeat the reducer m times for m inputs. The following error may be thrown as a result of not repeating the reducer: