Scribd
Upload
50 views

TSRD Java Homework Assignment

This document provides specifications for a homework assignment to develop a Java application that reads time series data from an input file, applies calculations to the data, and handles instrument price modifiers from a database. The application is to: 1) Read instrument price data from a simple input file format. 2) Query a database for instrument price modifiers. 3) Apply calculations such as mean, sum, or other statistics depending on the instrument. 4) Output the final calculated values to the console. Technical requirements include using Maven and a lightweight database with file-based storage. Design considerations focus on handling large data efficiently and scaling the solution.

Uploaded by

Rômero Ricardo
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
50 views

TSRD Java Homework Assignment

This document provides specifications for a homework assignment to develop a Java application that reads time series data from an input file, applies calculations to the data, and handles instrument price modifiers from a database. The application is to: 1) Read instrument price data from a simple input file format. 2) Query a database for instrument price modifiers. 3) Apply calculations such as mean, sum, or other statistics depending on the instrument. 4) Output the final calculated values to the console. Technical requirements include using Maven and a lightweight database with file-based storage. Design considerations focus on handling large data efficiently and scaling the solution.

Uploaded by

Rômero Ricardo
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Time Series fetching and calculations - Java

homework assignment task

Introduction
In TSRD we pull time series market data from various external providers and perform hundreds of calculations on the obtained data in order
to synthesize additional factors required by the downstream systems to compute the market risk levels.

Your task
In this excercise, we would like to emulate a simplified version of the scenario described above, by asking you to develop a simple
application that is going to read time series data from the specified input file and then applying several simple algorithms to the data received,

Input data
The input file format is very simple and it's basically a list of data records (one record - one input file line), where each record looks as
follows:

<INSTRUMENT_NAME>,<DATE>,<VALUE>

and:

INSTRUMENT_NAME - name of the financial instrument associated with a given data point; a real-life example would be e.g. USD/EUR
exchange rate or a company stock name etc.

DATE - a date on which a given instrument had a given value

VALUE - evaluation of a given instrument on a given date (e.g. price)

Example: USDPLEXRATE,12-Mar-2015,4.02

Calculations
Depending on the instrument, we would like to apply different algorithms to the received time series, namely:

INSTRUMENT1 – mean of all the values


INSTRUMENT2 – mean of all the values from November 2014
INSTRUMENT3 – any other statistical calculation that we can compute "on-the-fly" as we read the file (it's up to you)
any other instrument from the input file - sum of the newest 10 instrument values (in terms of the date).

The values above, however, are not just the values read from the input file as we would like to emulate the fact that in the real life there are
often multiple factors influencing calculations performed on the instrument prices.

So as part of your task we would like you to set up a database with only one table, called INSTRUMENT_PRICE_MODIFIER with the
following columns:

column meaning

ID primary key (just a db sequence number, no business meaning)

NAME instrument name (corresponding to the names from the input file)

MULTIPLIER double value that specifies the factor by which the original instrument value should be multiplied.

Moreover, we are only interested in business dates, i.e. Monday-Friday. Discard entries with non-business date.

To summarize then, in order to determine the final value of a given instrument price for a given date you should do the following:

1. read the line from the input file;


2. check if it's a record for a business date or not, if not - discard;
3. if it's a business date record - query the database to see if there is an entry for the <INSTRUMENT_NAME> you read in the 1st step;
4. if there is - multiply the original <VALUE> by the factor you found in the step 3;
5. if there is no entry - simply use the original <VALUE> from the file.

Please assume that the values in the INSTRUMENT_PRICE_MODIFIER table can change frequently independently of the process you're
implementing, but not more often than once every 5 seconds.

The following picture summarizes the task:


The final results can be printed out simply on the console, we don't want you to come up with any fancy UI or anything. Focus on the core of
the assignment.

Technical requirements
create the Maven (3+) project for your solution; we must be able to build your program using Maven;
use at least the JVM 1.7 version;
use a lightweight database (e.g. h2, Derby), memory or file-based - your choice, that doesn't require installation of any prerequisite
software and can be launched and torn down easily as part of your program run;
you’re allowed to use any Open Source libraries and frameworks you like, so e.g. you’re free to use Mockito, JUnit, Guava, Apache
Commons, Spring etc.
provide instructions on how to run your program (e.g. a java command with arguments allowing us to execute your program)
provide a description of your solution: important decisions made etc.

Guidelines and gotchas


model the problem using OO principles; think of what's likely to be a changing factor in this application and what's fairly constant;
design your solution accordingly;
test your work;
bear in mind that your program should also work for a file that has many gigabytes of data; it should be capable of handling it without
crashing along the way;
assume it's vitally important to calculate the abovementioned metrics as quickly as possible; your solution should scale effectively;
do not try to show off and don't over-engineere: your solution won't be evaluated on the basis of the number of frameworks used but
rather on the beauty of your design and the efficiency of the algorithms used;
send the source code back via email to us in a zip archive (but please don't include any executables or jars ! otherwise it will be
blocked by our corporate SPAM filter)

ABOVE SPECIFICATION IS AND ALWAYS WILL BE USED FOR CANDIDATE SKILLS EVALUATION ONLY.

You might also like