Tick History Request - Parametrize and Parallelize

Python Jupyter Notebook

February 09, 2022

Zoya Farberov

Developer Advocate

Last Updated: Jan 2026

Introduction

This article will discuss a quick and easy wrapper approach for running parallel Tick History requests. We are going to use TickHistoryTimeAndSalesExtractionRequest as an example request, while the analogous steps can be used with other Tick History and Datascope Select requests.

The companion example code is a Python Jupyter notebook* and is based on Tick History Python Code Samples*:

With this in mind, let us embark on a little coding voyage.

Approach in Brief

Authentication- token request
Parametrize request (TickHistoryTimeAndSalesRequest)
On Demand extraction request
- Define
- Extraction status polling
- Extraction notes retrieval
- Parallelize
Data retrieval and save to disk (includes AWS download)
- Define
- Parallelize

* see References section for the links to download example code

Import Required Modules and Define Required Parameters

This includes the definition of a "cookie-cutter" request:

Read Credentials from File

It is not essential how credentials are acquired, they can be entered as input pramterese or hard-coded, most importantly, that we have them at hand, for the required authentication step.

Authentication - Datascope Token Request

Parametrise Our Request

resulting in required requests to be ready to go:

Send On-demand Extraction Request - Define

We define a helper function to issue a request, that can consequently be issued in parallel

Send On-demand Extraction Request - Parallelize

This example code shows 2 requests issued in parallel. We can issue more requests in parallel, using the same pre-defined helper function. However, for TickHistoryTimeAndSalesRequest per Tick History REST API User Guide*, at the time of this writing, the number of concurrent processing requests is 2, same as for most other Tick History Templates. The maximum number of the requests that can be submitted is 50.

Therefore, per design, we should never submit more then 50 requests - maximum accepted. If we wish to only submit requests that will be immediately accepted for processing, we should submit 2 requests at a time. If we prefer to submit more and have some requests awaiting their turn, we can submit up to maximum allowed to be submitted, up to 50 requests.

if we are running in Jupyter notebook, event loop is already running and our tasks are executing in parallel, for example:

all that we need to do is await their completion: