Data analytics/visualization in Python

Data analytics/visualization in Python Note: This is a Test Task that will lead to the main project if the test is completed as per expectation. What follows are instructions for completing a test task for consideration as a Python-based data visualization, analytics and general scripting developer. When you are finished, please submit a pull request with your solution, including any artifacts, instructions, or clarifications. Deploy your solution in your preferred fashion and include a URL to the GUI in your submission. We will review your submission and get back to you within a few days! Background In order to monitor and improve our algorithms, we rely on various forms of analytics and analysis tools to understand the results of our trading decisions. We are striving to improve this infrastructure and hopefully you can help! Internally, we maintain various sets of data stores. One of them is called Bashoop which serves as a high-capacity time series data store, and is relied upon as an intermediary for analysis that relies upon time series data. See below for instructions on how to access a dummy instance of Bashoop created for this task. You will need to know some basics of market data and trading concepts. Also, a basic understanding of time series data. These are good starting points: Order Books ([login to view URL]) Bid/Ask ([login to view URL]–ask_spread) The Task The overall task will involve extracting market data for various trading instruments (represented by unique symbols), applying some transformations and statistical functions, and displaying the results in a report. 1. Obtain the following QUOTES series from Bashoop: TYQUOTES, FVQUOTES The quotes data is stored as a BBO. BBO stands for best bid/offer, which represents the best bid and ask (ask and offer are interchangable terms) in the market for a given instrument at any given time. At each timestamp it contains, in the following order: bid price, bid quantity, ask price, ask quantity 2. Calculate a VwMpt at each timestamp t for each instrument's QUOTES series VwMpt_t = bid_price_t + (ask_price_t-bid_price_t) * (bid_qty_t / (bid_qty_t+ask_qty_t)) 3. Calculate a rolling correlation rho of VwMpt returns (first difference, i.e. derivative) sampled at 1 minute intervals with a lookback of 100 samples 4. On a webpage, display a plot of VwMpt substracted from the first sample (so the time series starts at 0) of both series on the same line chart, and below it plot the rho lined up along the same time axis. You may use a library like Flot or something similar. Bashoop The Bashoop instance for this task can be reached at: [login to view URL] Bashoop data is stored, and can be retrieved, as flat files. This is possible via a RESTful interface Each Bashoop file is called a series and has a unique identifer (the series_id) The first column of each series (file) is a monotonically increasing timestamp stored as microseconds since epoch. This is then followed by one or more columns of values recorded at the given row's timestamp. To download a series with a series_id of FVQUOTES from a Bashoop instance located at [login to view URL] you can: curl -s -d "series_id=FVQUOTES" -d "start_ts=0" -d "end_ts=1451606400000000" -d "compress=0" [login to view URL] -X GET where start_ts and end_ts are micros send epoch. In this case we want the full history. Hints You can assume that BBO data remains unchanged between each update (each row) Assessment Criteria Code quality and documentation Design decisions Solution Usability