Description

This programs query Coinpaprika API (https://api-pro.coinpaprika.com/v1/tickers) every 10 seconds and saves the results to a csv file. It also prints the results to the console.

Columns

• ticker_id - Coinpaprika API_ID
• last_update - last_updated field from the API (date/time format)
• last_update_timestamp - same as above but as a timestamp
• now - current date/time of loop run from local machine
• now_timestamp - same as above but as a timestamp
• price - quotes.USD.Price field from the API
• price_changed - true if the price changed since the last run, false otherwise

Running

• Go - preferred
• Shell script - alternative

Go

• all parameters besides key are optional

• default output file is result.csv

• go run main.go -key=your_api_key - this will run the program, print the results to the console and save them to a file called result.csv

• go run main.go -key=your_api_key -file=filename.csv - if you wish to change the output location

• go run main.go -key=your_api_key -file=filename.csv -ticker=eth-ethereum - if you wish to change the ticker

• 

SHELL

If you don't have Go installed, you can use the provided main.sh file to run the shell script, which produces similar results.

• edit main.sh and add your API key
• sh main.sh | tee result.csv

Analysis

1. clickhouse local is used for analysis, just because it's our database of choice at the moment. You can work with Python/Pandas, DuckDB or any other database of your choice, using the CSV file as input.
2. You should run the program for a while (hour+) and use default output file result.csv
3. Example file with 1 hour of data (UTC: 2023-08-02 10:26 - 2023-08-02 11:26) is provided in the repository result.csv

Setup

curl https://clickhouse.com/ | sh - this will install a local version of Clickhouse in current directory

Queries

1. query_1.sql - show max, avg, stdDev time between price changes based on local time now_timestamp (in seconds)
   • ./clickhouse local < query_1.sql
2. query_2.sql - show max, avg, stdDev time between price changes based on Paprika time last_update_timestamp (in seconds)
   • ./clickhouse local < query_2.sql

Running these queries on the provided result.csv file should give you the following results: (with the following average response time around 90 seconds)