User manual

api-24sea-ai is an extension package for API-24SEA that provides tools to interact with the AI models hosted on the 24SEA platform.

Note

The following example shows the classical usage of the datasignals module. The first step is package installation, and after that the example code cells show how to:

• Import the package and other useful libraries.

• Load the environment variables from a .env file. This step is optional, since if any of the following names for user and password in the system, the package will authenticate automatically.

• API_24SEA_USERNAME, 24SEA_API_USERNAME, TWOFOURSEA_API_USERNAME, API_TWOFOURSEA_USERNAME for the username.

• API_24SEA_PASSWORD, 24SEA_API_PASSWORD, TWOFOURSEA_API_PASSWORD, API_TWOFOURSEA_PASSWORD for the password.

• After that, API can be instantiated and authenticated.

• Finally, the user can get predictions from the API. The AsyncAPI class will authenticate lazily if the environment variables are loaded, or the user can authenticate manually before performing the data retrieval.

Installation

The package supports Python [3.8:3.14). To install it, run the following command in your terminal:

pip install api-24sea-ai

or, being an extension package of api-24sea, you can install it as:

POWERSHELL

pip install api-24sea[ai]

BASH

pip install 'api-24sea[ai]'

Importing the package

# %%
# **Package Imports**
# - From the Python Standard Library
import logging
import os
import sys

# - From third party libraries
import pandas as pd
import dotenv  # <-- Not necessary to api-24sea-ai per se, but useful for
               #     loading environment variables. Install it with
               #     `pip install python-dotenv`

# - Local imports
from api_24sea.ai.version import __version__, parse_version

# %%
# **Package Versions**
print("Working Folder: ", os.getcwd())
print(f"Python Version: {sys.version}")
print(f"Pandas Version: {pd.__version__}")
print(f"Package {parse_version(__version__)}")
# **Notebook Configuration**
logging.basicConfig(level=logging.INFO)

Setting up the environment variables (optional)

This step assumes that you have a file structure similar to the following one:

.
├── env
│   └── .env
├── notebooks
│   └── example.ipynb
└── requirements.txt

The .env file should look like this:

API_24SEA_USERNAME=your_username
API_24SEA_PASSWORD=your_password

With this in mind, the following code snippet shows how to load the environment variables from the .env file:

# %%
# **Load Environment Variables from .env File**
_ = dotenv.load_dotenv("../env/.env")
if _:
    print("Environment Variables Loaded Successfully")
    print(os.getenv("API_24SEA_USERNAME"))
    # print(os.getenv("API_24SEA_PASSWORD"))
else:
    raise Exception("Environment Variables Not Loaded")

Performing AI model predictions

After loading the environment variables and authenticating with the API, the user can perform AI predictions from 24SEA API AI endpoints.

The predictions are returned as a DataFrame. All the metrics predicted from the requested models are stored in separate columns, and the timestamps are set as the index of the DataFrame.

The prediction is done by specifying the sites or the locations or both, the metrics, and timestamps.

• Sites: Case insensitive, it can either match site or site_id. It is an optional parameter.

• Locations: Case insensitive, it can either match location or location_id. It is an optional parameter.

• Models: Case insensitive, it can be a partial match of the model name. It is a required parameter.

• Timestamps: Timezone-aware datetime, strings in ISO 8601 format, or shorthand strings compatible with the shorthand_datetime package.

# %%
# **Authenticating AI Client and Viewing Models Overview**
from api_24sea.ai.core import AsyncAPI
api = AsyncAPI()
api.models_overview  # <-- View the models overview (performs lazy authentication)

# %%
# **Making Predictions**
api.get_predictions(
    sites="WF",
    locations="A01",
    model="mean_WF_A01_some_ai_model",
    start_timestamp="2020-03-01",
    end_timestamp="2020-06-01",
).head()