Python SDKs

​

Overview

dappier-py provides a straightforward way to interact with Dappier’s API’s, which allows for real-time data search on the internet and other datamodels from the marketplace. The library is designed to be easy to use and integrate into existing Python projects.

​

Installation

To install the package, run:

pip install dappier

Alternatively, you can clone the repository and install the dependencies:

git clone https://github.com/DappierAI/dappier-py
cd dappier-py
pip install -r requirements.txt

​

Initialization

You can get your API key from your Dappier account.

from dappier.dappier import DappierApp

app = DappierApp(api_key='your_api_key')

​

Real-Time Search

You can perform a real-time search by providing a query. This will search for real-time data related to your query.

result = app.realtime_search_api("When is the next election?")
print(result.response['response']["results"])

​

AI Recommendations

The AI Recommendations feature allows you to query for articles and other content using a specific data model. You can pick a specific datamodel from marketplace

​

Default Options:

ai_result = app.ai_recommendations(query="latest tech news", datamodel_id="dm_02hr75e8ate6adr15hjrf3ikol")
print(ai_result.results)

​

Custom Options:

You can pass custom parameters such as similarity_top_k, ref and num_articles_ref:

ai_custom_result = app.ai_recommendations(
    query="latest tech news",
    datamodel_id="dm_02hr75e8ate6adr15hjrf3ikol",
    similarity_top_k=5,
    ref="techcrunch.com",
    num_articles_ref=2
)
print(ai_custom_result.results)

​

Parameters

​

query (string):

• A natural language query or URL.
• If a URL is passed, the AI analyzes the page, creates a summary, and performs a semantic search query based on the content.

​

similarity_top_k (integer):

• The number of articles to return (default is 9).

​

ref (string):

• The domain of the site from which the recommendations should come.
• Example: techcrunch.com.

​

num_articles_ref (integer):

• Specifies how many articles should be guaranteed to match the domain specified in ref.
• Use this to ensure a set number of articles from the desired domain appear in the results.

​

search_algorithm (string):

• Options: "most_recent" or "semantic".
• "semantic" (default): Contextual matching of the query to retrieve articles.
• "most_recent": Retrieves articles sorted by the most recent publication date.

Checkout example.py in this repository for a working example.