Geospatial Data

Redis handles geographic coordinates using Sorted Sets. Longitude and latitude are encoded into a 52-bit GeoHash value, which serves as the score, while the location identifier acts as the member. This enables efficient spatial queries and distance computations.

Key Commands

• GEOADD: Stores geographic coordinates into a key.

GEOADD cities -122.4194 37.7749 "SanFrancisco" -118.2437 34.0522 "LosAngeles"

• GEODIST: Computes the distance between two members.

GEODIST cities "SanFrancisco" "LosAngeles" mi

• GEOPOS: Retrieves the coordinates for specified members.

GEOPOS cities "SanFrancisco" "LosAngeles"

• GEORADIUS: Queries members within a radius from a central coordinate.

GEORADIUS cities -122.0 37.0 100 mi

• GEORADIUSBYMEMBER: Queries members within a radius from an existing member.

GEORADIUSBYMEMBER cities "SanFrancisco" 100 mi

Practical Applications

• Proximity detection for ride-sharing or dating apps to find nearby users.
• Location-based services, such as locating nearby stores or points of interest on a map.

HyperLogLog

HyperLogLog provides a probabilistic approach to cardinality estimation. Instead of storing every unique element, it uses hashing to estimate the number of distinct items. This guarantees a constant memory footprint of roughly 12KB, regardless of the number of elements counted, at the cost of a small standard error rate.

Key Commands

• PFADD: Inserts elements into the HyperLogLog.

PFADD unique_visits userA userB userC

• PFCOUNT: Returns the approximate cardinality of the observed set.

PFCOUNT unique_visits

• PFMERGE: Combines multiple HyperLogLogs into a single one.

PFMERGE total_visits q1_visits q2_visits

Practical Applications

• Calculating unique visitor counts across distributed systems without high memory overhead.
• Tracking Daily Active Users (DAU) or Monthly Active Users (MAU) for platforms.
• Estimating unique page views or event occurrences in high-traffic scenarios.

Bitmaps

Bitmaps are not an intrinsic data type but a set of bit-oriented operations applied to the String data type. They allow setting or retrieving bits at specific offsets, enabling highly memory-efficient representation of binary states.

Key Commands

• SETBIT: Assigns a bit value (0 or 1) at a given offset.

SETBIT user:login:status 1001 1

• GETBIT: Fetches the bit value at a specific offset.

GETBIT user:login:status 1001

• BITCOUNT: Counts the number of set bits (1s) in a range.

BITCOUNT user:login:status 0 100

Practical Applications

• Tracking boolean states like user online/offline presence, where the offset represents the user ID.
• Feature flag management, toggling features on or off for specific users.
• Performing fast aggregate calculations, such as counting daily logins.
• Serving as the underlying mechanism for Bloom filters to check for potential set membership.