In this personal project, I use R programming to obtain cryptocurrency price data from Coinbase. Coinbase is a well-known cryptocurrency exchange for buying, selling and trading cryptocurrencies.

Reference: https://www.cryptodatadownload.com/blog/how-to-download-coinbase-price-data-R.html

## Topics

 

## Cryptodatadownload Blog Post On Obtaining Data Into R

Blog Post: https://www.cryptodatadownload.com/blog/how-to-download-coinbase-price-data-R.html

From the blog post, a script is provided where there is a function which takes a crypto pair and filename as inputs.

The Coinbase API is called and a .csv file is created based on the filename input. The later sections will build upon this script.

The script below is better if it is run in RStudio.

# Reference: https://www.cryptodatadownload.com/blog/how-to-download-coinbase-price-data-R.html

#Import libraries needed
# Use install.packages("glue") if needed
library(jsonlite)
library(glue)

# Create a function to retrieve daily data
retreive_daily_data <- function(pair) {
  url = glue("https://api.pro.coinbase.com/products/{pair}/candles?granularity=86400")
  columnNames <- c('unix', 'low', 'high', 'open', 'close', glue('{pair} volume'))
  mydata <- fromJSON(url)
  df <- as.data.frame(mydata)
  colnames(df) <- columnNames  # rename the columns
  
  write.csv(df, file = filename)
}

newPair <- "BTC-USD"
fileName <- glue("dailyData{newPair}.csv")
runFunc <- retreive_daily_data(newPair, filename = fileName)
runFunc

 

## Creating A Function For Producing Candlestick Charts From Coinbase Data

What if I do not want to write a .csv file and want to work with the data right away?

I have modified the blog post script. It is a function which takes in a crypto trading pair as the only input. The function extracts crypto data from Coinbase’s API.

Column names are renamed and UNIX times are converted into a datetime.

Candlestick price charts are produced as the output with the use of plotly. Plotly is nice as it is interactive compared to ggplot2.

 

About Granularity

Reference: https://docs.pro.coinbase.com/?r=1#get-historic-rates\

At first, I was confused about the granularity part. I asked myself why it was 86400. This 86400 represents 86400 seconds in a day. Setting the granularity to 86400 means that the obtained data would be daily price data.

Setting the granularity to 60 means price data every minute. Having the granularity at 300 would be every 5 minutes price data.

For the candlestick plotting, I am doing daily data. The granularity is at 86400.

library(tidyr)
library(dplyr)
library(ggplot2)
library(plotly)
library(jsonlite)
library(glue)
# Function Creation:
coinbase_candlestick_plot <- function(pair){
  
  # Obtain url given the crypto pair
  url = glue("https://api.pro.coinbase.com/products/{pair}/candles?granularity=86400")
  crypto_data <- fromJSON(url)
  df <- as.data.frame(crypto_data)
  
  # Renaming column names:
  columnNames <- c('unix', 'low', 'high', 'open', 'close', glue('{pair} volume'))
  colnames(df) <- columnNames
  
  # Change unix epoch time into date, change column names.
  df$unix <- as.POSIXct(df$unix, origin = '1970-01-01', tz = 'GMT')
  
  colnames(df) <- c('Date', 'Low', 'High', 'Open', 'Close', 'Volume')
  
  # Plotly Candlestick Chart
  df %>% plot_ly(x = ~Date, type="candlestick",
                                 open = ~Open, close = ~Close,
                                 high = ~High, low = ~Low) %>% 
  layout(title = paste("Basic Candlestick Chart Of", pair, "\n"),
        xaxis = list(rangeslider = list(visible = F)),
        yaxis = list(title = paste0(pair, ' Price \n')))
}

 

Some Function Calls

Not that the function for the plot is created. Some function calls can be performed to produce some candlestick price charts on various crypto trading pairs.

There are many crypto trading pairs out there on Coinbase. You can use Coinmarketcap’s website or Coingecko to look at available crypto trading pairs.

 

# Function call #1 (Ethereum priced in USDT/Tether):

coinbase_candlestick_plot("ETH-USDT")

 

# Function call #2 (Solana/SOL priced in USDT/Tether) :

coinbase_candlestick_plot("SOL-USDT")

 

# Function call #3 (Polkadot/DOT priced in USDT/Tether) :

coinbase_candlestick_plot("DOT-USDT")