The

load_assets_from_dbt_cloud_job

does not bring in the automatic dependency as the

load_assets_from_dbt_project

function does. Is this intended?

Hey Shane, I’d check the exact

AssetKey

for the

open_weather

dbt source that’s ingested. The `AssetKey`’s have to be exactly the same for the dependency to be understood properly. cc @claire, who is working on utilities to make this more ergonomic

I don't understand why I'm struggling with this so much. It feels not intuitive and I've went over the documentation a ton. Here's what my

@asset

looks like:

@asset(
    group_name="weather", 
    compute_kind="API", 
    io_manager_key="snowflake_io_manager",
    non_argument_deps="open_weather"
)
def open_weather() -> pd.DataFrame:
    cities_data = []
    for city in cities:
        url = f"<https://api.openweathermap.org/data/2.5/weather?q={city}&units=imperial&appid=apikey>"
        weather = requests.get(url).json()
        cities_data.append(weather)
        df = pd.json_normalize(cities_data)
        df = df.drop(columns=["weather"])

        df["dt"] = pd.to_datetime(df["dt"], unit="s").dt.date # Convert to date which will be used to partition by
    return df

Here's what my `sources.yml`file looks like:

version: 2

sources:
  - name: weather
    database: dev  
    schema: skalepp  
    tables:
      - name: open_weather

And this is what it loads. Any help would be appreciated.

Hi Shane, I think the

non_argument_deps

arg accepts a set instead of a string. In this case, it's interpreting your string as a set. We really should raise an error for this, but I think you can define your asset instead with:

@asset(non_argument_deps={AssetKey("open_weather")})
def open_weather():
    ...