twitter

etherscan

website

Blockchains (L2)

Unichain

# Unichain Active addresses (monthly) Methodology

## 1. Overview

Monthly active addresses measures the number of unique addresses that initiated at least one transaction on Unichain within a 30-day rolling window. This metric provides insights into user engagement and network activity by tracking distinct transaction senders over monthly periods. The calculation uses HyperLogLog (HLL) sketching for efficient approximate counting of large-scale distinct addresses across rolling time windows.

## 2. Data sources

- **Unichain transaction records**: Raw blockchain transaction data including transaction hashes, block timestamps, and sender addresses (`from_address`)
- **Unichain block data**: Block timestamp information for temporal aggregation and time-series construction

## 3. Step-by-step calculation

### Component A: Daily active user extraction
- **Purpose**: Identifies unique transaction senders per day as the foundation for monthly calculations
- **Logic**: Processes raw Unichain transaction records to extract distinct sender addresses (`from_address`) for each day. Timestamps are truncated to daily granularity using `timestamp_trunc` function. Each transaction sender is classified as user type "Transaction sender" with chain metadata including market sector "blockchains_l2" and business segment "blockchain". The daily distinct addresses become the input for monthly rolling window calculations.
- **Key fields**: `block_timestamp` (truncated to day), `user_address` (from `from_address`), `chain`, `user_type`

### Component B: HyperLogLog initialization for daily users
- **Purpose**: Converts daily user addresses into HLL sketches for efficient approximate distinct counting across large datasets
- **Logic**: Daily distinct user addresses are processed through HLL initialization using `hll_count.init` with precision parameter 15. This creates compact sketches that can be efficiently merged across time windows while maintaining approximate distinct count accuracy. Each daily HLL sketch represents the unique addresses active on that specific day.
- **Key fields**: `hll` (HyperLogLog sketch), `timestamp`, `chain_id`, `user_type`, `product_id`

### Component C: Rolling window construction
- **Purpose**: Creates 30-day rolling windows for each date to calculate monthly active addresses
- **Logic**: Uses a datespine approach covering the full time range from minimum timestamp to current date minus one day. For each date in the spine, generates a 30-day lookback window (current day plus previous 29 days) using `generate_timestamp_array`. Creates join conditions between spine dates and their corresponding 30-day windows. This ensures complete temporal coverage without gaps and enables consistent rolling window calculations.
- **Key fields**: `timestamp` (window end date), `joinday` (individual days within each 30-day window)

### Component D: HLL sketch aggregation within windows
- **Purpose**: Merges daily HLL sketches within each 30-day window to calculate approximate distinct user counts
- **Logic**: For each 30-day window, collects all daily HLL sketches falling within that period. Uses `hll_count.merge_partial` to combine sketches from the same window, creating a unified sketch representing all unique addresses active during those 30 days. This approach efficiently handles large-scale distinct counting while maintaining reasonable accuracy for monthly user metrics.
- **Key fields**: Merged `hll` sketches per 30-day window, window metadata

### Component E: Dimensional breakdown creation
- **Purpose**: Generates multiple dimensional views of monthly active addresses for different analytical perspectives
- **Logic**: Creates four separate dimensional breakdowns from the windowed HLL data: project-level aggregation (dimensions = "project_dbt_tag") combines all chains and user types into protocol-level metrics; chain-level aggregation (dimensions = "chain") breaks down by individual blockchain while nullifying other dimensions; user type aggregation (dimensions = "user_type") segments by user classification; product-level aggregation (dimensions = "product_id") breaks down by specific products. Each breakdown applies `hll_count.extract` to convert merged HLL sketches into final distinct user counts.
- **Key fields**: `dimensions`, `user_mau`, `chain_id`, `user_type`, `product_id`

### Final aggregation
- **Combining components**: All dimensional breakdowns are combined using UNION ALL operations to create a comprehensive dataset with multiple analytical perspectives. The target model filters for only "chain" and "project_dbt_tag" dimensions to focus on the most relevant breakdowns for active addresses reporting.
- **Currency conversion**: No currency conversion required as this metric counts addresses rather than monetary values
- **Time aggregation**: Rolling 30-day windows are calculated continuously, with each date representing the month ending on that date. Monthly boundaries are implicit through the 30-day lookback period rather than calendar month alignment.
- **Final calculation**: The user_mau field from monthly active user calculations is renamed to `active_addresses_monthly` to reflect the specific metric being measured. Standard metric finalization applies zero-filling for missing periods and ensures consistent time-series output format.

*This methodology was automatically generated by Token Terminal's methodology analyzer.*

Unique addresses active over a rolling 30-day window.

UnichainUNI

Active addresses (monthly)

Active addresses (monthly)

Competitive landscape

Chain breakdown