README.md

Financial Data Visualization using doViz (Beta1.0)

doViz (dodo🦤Visualizer) is an open source financial data visualization tool from tradingDodo.

• doViz can represent any sequential/streaming data with or without timestamp
• doViz can be used to analyze local time-series CSV files by analyzing them sequentially
• currently in doViz beta1.0 we are starting with visualizing financial data
• Stuff you can currently do with doViz:
  • Read and parse price files, multiple file formats and file kinds like OHLC data, Volume data, Short Interest data can be combined.
  • Stock Price, Volume, OHLC, can be visualized for all dates available.
  • Represent Portfolios in terms of multiple attributes (Pie chart of stocks bought, profit made, etc).
  • Represent volume as heatmap for n days and visualize sequentially.

Installation

Currently we are on Beta, hence you can clone our repo and start using doviz. We'll be releasing API's for realtime data visualization pretty soon. We're planning to launch soon (by April 2021).

Using doviz for visualizing historical price data

The following content has information on how to use the basic version of doviz. In this beta version of doviz, you can visualize your data you have as csv files.

We first need to import doviz

from doviz import pricedatalayer as player

We can then define a historical price layer for a stock and call getPrice() everytime to get the next recorded price of the stock.

ph = player.priceHistorical("AAPL",20)
for tstamp in range(5):
    print("price %s" %ph.getPrice())

You should be seeing something like: price ['12/30/2020', 133.72, 96452120] price ['12/31/2020', 132.69, 99116590] price ['01/04/2021', 129.41, 143301900] price ['01/05/2021', 131.01, 97664900] price ['01/06/2021', 126.6, 155088000]

NOTE: The priceHistorical module makes us of the folder tickerdata, the prices of only those stocks are usable by the module whose price data is available inside the tickerdata folder. It returns the timestamp followed by the price and volume. If volume isn't available it will not be returned.

Analyzing Historical daily free data from NASDAQ

Historical daily data from NASDAQ work perfectly with the historical price module. Since the authors of the repo used NASDAQ daily data for the work, they made sure any downloaded data from their page works. An example of data is available in the link below. https://www.nasdaq.com/market-activity/stocks/aapl/historical Additional stock price data available. You can download all stock data you are interested in, add their CSV files inside the tickerdata folder, and then run the script below to see which Tickers you collected. Some, ticker files collected have been included in the repo, but the data will not be the latest data.

#FUNCTION to get all TICKER's as a list from the tickerdata folder 
import glob
def getalltickers():
    alltickers = []
    for el in glob.glob("tickerdata/*.csv"):
        tmp = el.split("\\")[1]
        finx= tmp.split(".")[0]
        alltickers.append(finx)
    return alltickers

print(getalltickers())

Defining a priceDataLayer from a date

in player.priceDataLayer("01/12/2018",0.5) priceDataLayer stores information of multiple tickers, and hence we need to define these thigs here:

• date_of_visualization_start = (%m/%d/%Y) eg. "01/12/2018", please note that if a trading data on a date is not available, mostly because its a Saturday/Sunday, you'll get an error saying "this day is untraded". In that case you can add or subtract days till you don't get the error anymore.
• refresh_seconds = 0.5, which means after every 0.5seconds the prices of all the tickers will be refresged

in pdl.addTickers(["UAL","DAL","JBLU"],[-1])

• tickerList = the tickers you want to visualize
• priceContext = every time the new price is pulled, the ratio of the new to old price is stored as an element in the price-vector

Adding tickers to priceDataLayer

In the example below we add "UAL","DAL","JBLU" as tickers and we also add priceContext of 1

priceDataLayer.visualize

visualize is the primary function inside priceDataLayer. It has two modes,

• allplots : Shows (price, price velcity, volume bars, change heatmaps); Detached plots, i.e. if you enter 5 tickers you'll get 5 plots.
• priceonly : Shows (price, price velocity) for all tickers in a single plot. (With time we'll be adding more kinds of visualizations)

Visualizing prices of MSFT, AAPL, FB since 01/12/2018

• We define a priceDataLayer starting from 01/12/2018
• Add tickers MSFT, AAPL, FB (if any ticker isnt available on the day, it won't be displayed)
• (Optional) setting dark mode to True.
• We visualize with mode 'priceonly'

pdl = player.priceDataLayer("01/10/2019",1)
pdl.addTickers(["MSFT","FB","AAPL"],[-1]) 
pdl.setDarkMode(True)
pdl.visualize("priceonly")

The above code should start a window, which refreshes after every 1 second. The end result should look like this:

Now, lets try to see what happened to Gamestop this week? 01/31/2021 Unfortunately we don't have the second by second data for the day of the spike, but regardless, we can see the crazy activity over the last week, and the spike of 300%.

pdl = player.priceDataLayer("12/21/2020",0.5)
pdl.setDarkMode(False)
pdl.addTickers(["GME","AMC","INTC"],[-1]) 
pdl.visualize("priceonly")

Now, lets look at the volume data as well, by selecting parameter "allplots" for the function visualize.

pdl = player.priceDataLayer("10/21/2020",0.1)
pdl.setDarkMode(True)
pdl.addTickers(["GME","AMC"],[-1]) 
pdl.visualize("allplots")

Creating a dashboard for analysis of price data using doviz and ipywidgets

Here we have tries creating a simple dashboard, that retrieves the stock data ticker from your tickerdata folder and then creates a drop down menu Run the code below to get to the dashboard

from ipywidgets import interact, widgets
from IPython.display import display
import glob
from doviz import pricedatalayer as priceDataLayer

#function to retrieve all available tickers, before populating drop down menu
def getalltickers():
    alltickers = []
    for el in glob.glob("tickerdata/*.csv"):
        tmp = el.split("\\")[1]
        finx= tmp.split(".")[0]
        alltickers.append(finx)
    return alltickers

global_variables = list(globals().keys())
allops = getalltickers()

global_vars = widgets.SelectMultiple(
    options=allops)

butt = widgets.Button(description='Select Tickers!')
outt = widgets.Output()

text = widgets.Text(
    value='last',
    placeholder='Enter Simulation start date!',
    description='Enter Date to start Simulation!:',
    disabled=False
)
selectedTickers = None

def assignselected(stickers):
    print("Tickers to analyze:",stickers)
    global selectedTickers
    selectedTickers = stickers

def on_butt_clicked(b):
    with outt:
        clear_output()
        selectedTickers = list(global_vars.value)
        assignselected(selectedTickers)
        
def callback(wdgt):
    global selectedTickers
    datex = wdgt.value
    pdl = priceDataLayer.priceDataLayer(datex,1)
    pdl.setDarkMode(False)
    pdl.addTickers(selectedTickers,[-1]) #tickerpricelayerdays
    pdl.visualize("all")

text.on_submit(callback)
butt.on_click(on_butt_clicked)
widgets.VBox([global_vars,butt,outt,text])

This should give you a menu like this when you run the code. You can try creating a different dashboard Alternatively, the module for button and menu based running is being added to doviz for more flexibility and easier usage.