Dim Labels

-- Track CEX inflows and outflows
WITH cex_addresses AS (
    SELECT DISTINCT address
    FROM <blockchain_name>.core.dim_labels
    WHERE label_type = 'cex'
        AND label_subtype IN ('hot_wallet', 'deposit')
)
SELECT
    DATE_TRUNC('day', t.block_timestamp) AS day,
    CASE
        WHEN t.to_address IN (SELECT address FROM cex_addresses) THEN 'CEX Inflow'
        WHEN t.from_address IN (SELECT address FROM cex_addresses) THEN 'CEX Outflow'
    END AS flow_direction,
    COUNT(*) AS transaction_count,
    SUM(t.value) AS total_native_value,
    COUNT(DISTINCT t.from_address) AS unique_users
FROM <blockchain_name>.fact.fact_transactions t
WHERE (t.to_address IN (SELECT address FROM cex_addresses)
       OR t.from_address IN (SELECT address FROM cex_addresses))
    AND t.block_timestamp >= CURRENT_DATE - 30
    AND t.value > 0
GROUP BY 1, 2
ORDER BY 1 DESC, 2;

-- Top DeFi protocols by unique users
SELECT
    l.project_name,
    l.label_type,
    COUNT(DISTINCT t.from_address) AS unique_users,
    COUNT(*) AS total_interactions,
    SUM(t.tx_fee) AS total_fees_paid
FROM <blockchain_name>.fact.fact_transactions t
JOIN <blockchain_name>.core.dim_labels l ON t.to_address = l.address
WHERE l.label_type IN ('defi', 'dex')
    AND t.block_timestamp >= CURRENT_DATE - 7
    AND t.tx_status = 'SUCCESS'
GROUP BY 1, 2
ORDER BY 3 DESC
LIMIT 20;

-- Monitor large holder activities
WITH whale_activity AS (
    SELECT
        l.address,
        l.address_name,
        l.label_type,
        COUNT(*) AS tx_count,
        SUM(CASE WHEN t.from_address = l.address THEN 1 ELSE 0 END) AS outgoing_tx,
        SUM(CASE WHEN t.to_address = l.address THEN 1 ELSE 0 END) AS incoming_tx,
        SUM(CASE WHEN t.from_address = l.address THEN t.value ELSE 0 END) AS value_sent,
        SUM(CASE WHEN t.to_address = l.address THEN t.value ELSE 0 END) AS value_received
    FROM <blockchain_name>.core.dim_labels l
    JOIN <blockchain_name>.fact.fact_transactions t
        ON l.address IN (t.from_address, t.to_address)
    WHERE l.label_type IN ('whale', 'institution', 'fund')
        AND t.block_timestamp >= CURRENT_DATE - 1
    GROUP BY 1, 2, 3
)
SELECT
    *,
    value_received - value_sent AS net_flow
FROM whale_activity
WHERE tx_count > 10
ORDER BY ABS(net_flow) DESC;

-- Find addresses interacting with multiple protocols
WITH user_protocols AS (
    SELECT
        t.from_address AS user_address,
        l.project_name,
        l.label_type,
        COUNT(*) AS interactions
    FROM <blockchain_name>.fact.fact_transactions t
    JOIN <blockchain_name>.core.dim_labels l ON t.to_address = l.address
    WHERE l.label_type IN ('defi', 'dex', 'nft')
        AND t.block_timestamp >= CURRENT_DATE - 30
    GROUP BY 1, 2, 3
)
SELECT
    user_address,
    COUNT(DISTINCT project_name) AS protocols_used,
    SUM(interactions) AS total_interactions,
    ARRAY_AGG(DISTINCT project_name) AS protocol_list
FROM user_protocols
GROUP BY 1
HAVING COUNT(DISTINCT project_name) >= 5
ORDER BY 2 DESC
LIMIT 100;

-- Assess labeling coverage for active addresses
WITH active_addresses AS (
    SELECT DISTINCT address, address_count
    FROM (
        SELECT to_address AS address, COUNT(*) AS address_count
        FROM <blockchain_name>.fact.fact_transactions
        WHERE block_timestamp >= CURRENT_DATE - 7
        GROUP BY 1
        HAVING COUNT(*) > 100
    )
)
SELECT
    CASE WHEN l.address IS NOT NULL THEN 'Labeled' ELSE 'Unlabeled' END AS status,
    COUNT(*) AS address_count,
    SUM(a.address_count) AS total_transactions
FROM active_addresses a
LEFT JOIN <blockchain_name>.core.dim_labels l ON a.address = l.address
GROUP BY 1;

-- Check data latency
SELECT
    DATE_TRUNC('hour', block_timestamp) as block_hour,
    DATE_TRUNC('hour', inserted_timestamp) as insert_hour,
    AVG(TIMESTAMPDIFF('minute', block_timestamp, inserted_timestamp)) as avg_latency_minutes
FROM <blockchain_name>.core.fact_transactions
WHERE block_timestamp >= CURRENT_DATE - 1
GROUP BY 1, 2;
``` |
| MODIFIED_TIMESTAMP | TIMESTAMP_NTZ | UTC timestamp of the most recent update to this record.

**Format**: TIMESTAMP_NTZ

**Triggers for Updates**:
- Data corrections
- Enrichment additions
- Reprocessing for accuracy
- Schema migrations

**Monitoring Usage**:
```sql
-- Recently modified records
SELECT *
FROM <blockchain_name>.core.fact_transactions
WHERE modified_timestamp > inserted_timestamp
AND modified_timestamp >= CURRENT_DATE - 1;

-- Data quality tracking
SELECT
    DATE(modified_timestamp) as mod_date,
    COUNT(*) as records_updated,
    COUNT(DISTINCT block_number) as blocks_affected
FROM <blockchain_name>.core.fact_transactions
WHERE modified_timestamp > inserted_timestamp
GROUP BY 1
ORDER BY 1 DESC;
``` |