ntropy-sdk 3.6.1

SDK for the Ntropy API

Ntropy SDK

SDK for the Ntropy API

Installation:

$ pip install --upgrade 'ntropy-sdk[benchmark]'

Usage:

Programmatic usage for benchmarking:

Assuming you have a CSV file called testset.csv with the following fields set:

• iso_currency_code: The currency of the transaction
• amount: The amount (a positive number)
• entry_type: incoming for money coming into the account, outgoing for money going out of the account
• description: The transaction description string
• account_holder_id: A unique identifier for the account holder
• account_holder_type: A string indicating the type of the account of the transaction (business or consumer)
• (optional) correct_merchant: A string that should match the merchant output of the API. If you combine this with --ground-truth-merchant-field=correct_merchant you will be given an accuracy % for the API.
• (optional) correct_labels: The label that the transaction should be marked as. If you combine this with --ground-truth-label-field=correct_labels you will be given an F1 score for the API.

$ ntropy-benchmark --api-key=$NTROPY_API_KEY --in-csv-file=testset.csv --out-csv-file=enriched.csv --ground-truth-label-field=correct_labels

Using this as a python library:

import os
from ntropy_sdk.ntropy_sdk import SDK, Transaction

sdk = SDK(os.getenv("NTROPY_API_KEY"))

transaction = Transaction(
    amount=1.0,
    description="AMAZON WEB SERVICES AWS.AMAZON.CO WA Ref5543286P25S: Crd15",
    entry_type="outgoing",
    account_holder_id="1",
    account_holder_type="business",
    iso_currency_code="USD",
    country="US",
)

batch = sdk.enrich_batch([transaction])
enriched_list = batch.wait_with_progress()
print("BATCH", enriched_list.transactions[0].labels)

enriched = sdk.enrich(transaction)
print("REALTIME:", enriched.labels)

License:

Free software: MIT license