Cryptocurrencies-explorations

Cryptocurrencies explorations

Introduction

The main goal of this computational Markdown document is to provide some basic views and insights into the landscape of cryptocurrencies. The “landscape” we consider consists of price action and trading volume time series for cryptocurrencies found in Yahoo Finance.

In this document we compute and plot with Raku the statistics in [AA1].

Details on "running" the document

The Raku package used for data retrieval is "Data::Cryptocurrencies", [AAp1]. The JavaScript D3 plots are made via "JavaScript::D3", [AAp2].

This Markdown document is converted to its woven Markdown version with "Text::CodeProcessing", [AAp3]. The woven Markdown document is converted to an HTML document with "Markdown::Grammar", [AAp4].

Here is the corresponding shell command:

file-code-chunks-eval Cryptocurrencies-explorations.md &&
  from-markdown Cryptocurrencies-explorations_woven.md -t html -o Cryptocurrencies-explorations.html &&
  open Cryptocurrencies-explorations.html

Setup

Here we load the packages used below:

use Data::Cryptocurrencies;
use Data::Reshapers;
use Data::Summarizers;
use Text::Plot;
use JavaScript::D3;

Time series

Here we get Bitcoin (BTC) data from 1/1/2020 until now:

my %ccTS = cryptocurrency-data(<BTC ETH>, dates => (DateTime.new(2020,1,1,0,0,0), now), props => <DateTime Close>, format => 'hash'):cache-all;

say %ccTS.elems;

Here is a summary:

records-summary(%ccTS);

Here are D3.js plots:

my %ts4 = %ccTS.map({ $_.key => $_.value.map(-> %r { %( date => %r<DateTime>.Str.substr(0,10), value => %r<Close>.Numeric, group => $_.key ) }).grep({ $_.<value> ~~ Numeric }) });
say %ts4>>.elems;

deduce-type(%ts4<BTC>)

js-d3-date-list-plot(%ts4<BTC>, plot-label => 'BTC', width => 800, height => 400, format => 'html', div-id => 'BTC');

js-d3-date-list-plot(%ts4<ETH>, plot-label => 'ETH', width => 800, height => 400, format => 'html', div-id => 'ETH');

Pareto principle adherence

Get data for all cryptocurrencies:

my @dsData = cryptocurrency-data('all', dates => (now - 14 * 24 * 3600, now), props => ['Symbol', 'DateTime', 'Close', 'Volume'], format => 'dataset');
say "Dimensions : {dimensions(@dsData)}.";

Clean data and show summary:

@dsData = @dsData.grep({ $_<Close> ~~ Numeric });
#records-summary(@dsData);

Group by "Symbol" and find price- and volume totals per group:

my %groups = group-by(@dsData, "Symbol");

my %prices = %groups.map({ $_.key => $_.value.map(*<Close>).sum });
my %volumes = %groups.map({ $_.key => $_.value.map(*<Volume>).sum });

say %volumes.sort({ -$_.value }).head(5);

say text-pareto-principle-plot(%prices.values.List, title => 'Prices');
say text-pareto-principle-plot(%volumes.values.List, title => 'Volumes');

Here is the Pareto plot for closing prices:

my @cumSumPrices = produce(&[+], %prices.values.sort.reverse);
@cumSumPrices = @cumSumPrices X/ max(@cumSumPrices);
js-d3-list-plot(@cumSumPrices,
                plot-label => 'Pareto principle adherence for closing prices',
                width => 400, height => 300,
                format => 'html', div-id => 'pareto-prices'):grid-lines;

Here is the Pareto plot for trading volumes:

my @cumSumVolumes = produce(&[+], %volumes.values.sort.reverse);
@cumSumVolumes = @cumSumVolumes X/ max(@cumSumVolumes);
js-d3-list-plot(@cumSumVolumes,
                plot-label => 'Pareto principle adherence for trading volumes',
                width => 400, height => 300,
                format => 'html', div-id => 'pareto-volumes'):grid-lines;