Hex

https://hex.tech/docs
Admin
Hex is an AI Analytics Platform that provides a single, connected workspace for data practitioners...
Tokens:2,913
Snippets:41
Trust Score:10
Update:3 months ago
Show doc for...
Tokens:
Raw
### Aggregate Revenue by Customer Sector with Pandas

Source: https://hex.tech

Calculates total revenue and sector percentage share for each product line in Q3.

```python
# Aggregate revenue by product line + customer sector
df_sector = (
    df_q3.groupby(["product_line", "customer_sector"], as_index=False)
      .agg(total_revenue=("revenue_usd", "sum"))
)

# Compute share of each sector within a product line
df_sector["pct_of_product"] = (
    df_sector.groupby("product_line")["total_revenue"]
             .transform(lambda x: x / x.sum())
)

df_sector
```

--------------------------------

### Retrieve Q1-Q3 Revenue Data with SQL

Source: https://hex.tech

Fetches quarterly revenue data filtered by specific quarters from the galactic_sales.revenue table.

```sql
select quarter, product_line, region, customer_sector, revenue_usd
from galactic_sales.revenue
where quarter in ('Q1-2025', 'Q2-2025', 'Q3-2025')
```

--------------------------------

### Generate Synthetic Account-Level Data with Python

Source: https://hex.tech

Uses numpy to simulate account-level revenue and growth metrics based on Q3 sales data.

```python
import numpy as np

# Start from Q3 rows only
df_q3 = df[df["quarter"] == "Q3-2025"].copy()

# Repeat each row a few times to simulate multiple accounts per product/region/sector
REPEATS_PER_GROUP = 8  # tweak to control density
df_accounts = df_q3.loc[df_q3.index.repeat(REPEATS_PER_GROUP)].copy()

# Create a simple account id
df_accounts["account_id"] = (
    df_accounts["product_line"].str.replace(" ", "_").str.lower()
    + "-"
    + df_accounts.groupby("product_line").cumcount().astype(str)
)

# Split product-line revenue into pseudo account-level revenue with some noise
base_revenue = df_accounts["revenue_usd"] / REPEATS_PER_GROUP
df_accounts["account_revenue_usd"] = base_revenue * np.random.uniform(0.6, 1.4, size=len(df_accounts))

# Synthetic QoQ growth percentage for the scatter's Y-axis
df_accounts["qoq_revenue_growth_pct"] = np.random.uniform(-10.0, 30.0, size=len(df_accounts))

df_accounts
```

--------------------------------

### Aggregate Revenue by Region with Pandas

Source: https://hex.tech

Calculates total revenue and regional percentage share for each product line in Q3.

```python
import pandas as pd

# Aggregate revenue by product line + region
df_q3 = df[df["quarter"] == "Q3-2025"]
df_region = (
    df_q3.groupby(["product_line", "region"], as_index=False)
      .agg(total_revenue=("revenue_usd", "sum"))
)

# Compute share of each region within a product line
df_region["pct_of_product"] = (
    df_region.groupby("product_line")["total_revenue"]
             .transform(lambda x: x / x.sum())
)

df_region
```

--------------------------------

### Aggregate Revenue by Region and Sector with Pandas

Source: https://hex.tech

Calculates total revenue and percentage share per product line grouped by region or customer sector.

```python
import pandas as pd

# Aggregate revenue by product line + region
df_q3 = df[df["quarter"] == "Q3-2025"]
df_region = (
    df_q3.groupby(["product_line", "region"], as_index=False)
      .agg(total_revenue=("revenue_usd", "sum"))
)

# Compute share of each region within a product line
df_region["pct_of_product"] = (
    df_region.groupby("product_line")["total_revenue"]
             .transform(lambda x: x / x.sum())
)

df_region
```

```python
# Aggregate revenue by product line + customer sector
df_sector = (
    df_q3.groupby(["product_line", "customer_sector"], as_index=False)
      .agg(total_revenue=("revenue_usd", "sum"))
)

# Compute share of each sector within a product line
df_sector["pct_of_product"] = (
    df_sector.groupby("product_line")["total_revenue"]
             .transform(lambda x: x / x.sum())
)

df_sector
```

--------------------------------

### Define a Semantic Model in YAML

Source: https://hex.tech

Configure a semantic model to map warehouse tables to logical dimensions and metrics for agent queries.

```yaml
# Semantic model for NexaCorp's galactic sales data.
# This is the top-level model the agent queries by default.

model: galactic_sales

source:
  # Warehouse table backing this semantic model
  table: galactic_sales.revenue
  primary_key: revenue_id

dimensions:
  # Logical dimensions exposed to the agent
  - product_line        # defined in product_lines.yml
  - region              # defined in regions.yml
  - customer_sector     # defined in customer_sectors.yml
  - quarter

metrics:
  total_revenue_usd:
    label: Total revenue
    type: sum
    sql: revenue_usd
    format: currency

  avg_revenue_per_deal:
    label: Avg revenue per deal
    type: derived
    sql: total_revenue_usd / count_distinct(revenue_id)
    format: currency

  qoq_revenue_change:
    label: QoQ revenue change
    type: derived
    sql: total_revenue_usd.this_quarter - total_revenue_usd.prev_quarter
    format: currency
```
Tokens:
Raw