Report API: Sample scripts

Two Python scripts are available to help you work with the Report API: one that fetches real settlement data, and one that generates synthetic data for testing.

note

These scripts are starting-point examples, not production-ready code. They demonstrate the basic patterns for authentication and data retrieval, but do not include pagination, robust error handling, or retry logic.

Prerequisites

• Python 3.x

• For the fetch script: install the requests library:

  pip install requests

• For the fetch script: API credentials (see API keys)

Retrieve data from the Report API

This script authenticates, retrieves all your ledgers, then pulls both funds and fees entries for today from the Report API and saves them to a CSV file named by date (for example, 2026-03-03.csv).

Download fetch-report-api-data.py

Configure the script

Open the script and set the following variables before running it:

Variable	Description
USE_MIAMI_API	Set to True for accounting partners using the MIAMI API; False for merchants using their own API keys.
CLIENT_ID	Your client ID.
CLIENT_SECRET	Your client secret.
SUBSCRIPTION_KEY	Your Ocp-Apim-Subscription-Key. Required for merchants (USE_MIAMI_API = False).
MERCHANT_SERIAL_NUMBER	Your merchant serial number. Required for merchants (USE_MIAMI_API = False).

Run the script

python3 fetch-report-api-data.py

The settlement data is saved to a CSV file named after today's date (for example, 2026-03-03.csv).

CSV columns

The output file contains one row per ledger entry with these columns:

Column	API field
Reference (Vipps PSP)	pspReference
Time	time
Ledger Date	ledgerDate
Entry Type	entryType
Order ID/Reference (merchant)	reference
Currency	currency
Amount	amount
Balance Before	balanceBefore
Balance After	balanceAfter
Recipient Handle	recipientHandle
Message	message
Name	name
Masked Phone	maskedPhoneNo

Generate fake data

This script creates synthetic data that matches the shape of GET /report/v2/ledgers/{ledgerId}/{topic}/dates/{ledgerDate} responses. This is useful for testing data pipelines without connecting to the live API. It requires no external libraries and streams output to disk to avoid loading everything into memory at once.

Download generate-fake-report-api-data.py

Output files

The script writes one gzipped JSON file per day for each topic:

• out/funds/YYYY-MM-DD.json.gz
• out/fees/YYYY-MM-DD.json.gz

Arguments

Argument	Default	Description
--start	(required)	Start date in YYYY-MM-DD format.
--end	(required)	End date in YYYY-MM-DD format.
--payments	1000000	Total number of payments to distribute across the date range.
--out	./out	Output directory.
--ledger-id	302321	Ledger ID to use in the generated data.
--recipient-handle	NO:123455	Recipient handle to use in the generated data.
--currency	NOK	Currency code.
--net-settlement	(flag)	Simulate net settlement: adds daily fees-retained entries (negative in funds, positive in fees) and a payout-scheduled when the balance is positive. If omitted, gross settlement is simulated with a monthly fees-invoiced entry on the last day of each month.
--include-gdpr	0.10	Probability (0–1) that a transaction includes GDPR fields (name, message, maskedPhoneNo).
--seed	42	Random seed for reproducible output.
--tz	Z	Timestamp suffix.

Example

python generate-fake-report-api-data.py \
  --payments 1000000 \
  --start 2025-10-01 \
  --end   2025-10-31 \
  --out   ./out \
  --ledger-id 302321 \
  --recipient-handle NO:123455 \
  --currency NOK \
  --net-settlement \
  --include-gdpr 0.12 \
  --seed 42

This creates sample data for 1,000,000 payments across October 2025, distributed by weekday (Fridays and Saturdays are slightly busier). It produces realistic mixes of entry types including captures, refunds, disputes, and corrections for both funds and fees topics.