Analyze cryptocurrency market data
This tutorial is a step-by-step guide on how to analyze a time-series cryptocurrency dataset using TimescaleDB. The instructions in this tutorial were used to create this analysis of 4100+ cryptocurrencies.
The tutorial covers these steps:
- Design our database schema
- Create a dataset using publicly available cryptocurrency pricing data
- Load the dataset into TimescaleDB
- Query the data in TimescaleDB
You can skip ahead to the TimescaleDB portion if you would prefer not to run through the scripts to create your database schema or your dataset.
You can also download the resources for this tutorial:
Schema creation script:
Dataset creation script:
Dataset:
Crypto Currency Dataset September 2019
(Note that this data is from September 2019. Follow the steps in Section 2 of this tutorial if you require fresh data)
Prerequisites
To complete this tutorial, you need a cursory knowledge of the Structured Query Language (SQL). The tutorial walks you through each SQL command, but it is helpful if you’ve seen SQL before.
To start, install TimescaleDB. When your installation is complete, you can start ingesting or creating sample data.
This tutorial leads directly into a second tutorial that covers how Timescale can be used with Tableau to visualize time-series data.
Design the database schema
When you have a new database up and running, you need some data to insert into it. Before you get data for analysis, you need to define what kind of data you want to perform queries on.
In this analysis, we have two main goals:
- Explore the price of Bitcoin and Ethereum, expressed in different fiat currencies, over time.
- Explore the price of different cryptocurrencies, expressed in Bitcoin, over time.
Some questions you might want to ask:
- How has Bitcoin’s price in USD varied over time?
- How has Ethereum’s price in ZAR varied over time?
- How has Bitcoin’s trading volume in KRW increased or decreased over time?
- Which crypto-currency has the greatest trading volume in the last two weeks?
- Which day was Bitcoin most profitable?
- Which are the most profitable new coins from the past three months?
Understanding the questions you want to ask of the data helps to inform your schema definition.
These requirements lead us to four tables. We need three TimescaleDB hypertables, called btc_prices
, crypto_prices
, and eth_prices
, and one relational table, called currency_info
.
The btc_prices
and eth_prices
hypertables contain data about Bitcoin prices in 17 different fiat currencies since 2010. This is the Bitcoin table, but the Ethereum table is very similar:
Field | Description |
---|---|
time | The day-specific timestamp of the price records, with time given as the default 00:00:00+00 |
opening_price | The first price at which the coin was exchanged that day |
highest_price | The highest price at which the coin was exchanged that day |
lowest_price | The lowest price at which the coin was exchanged that day |
closing_price | The last price at which the coin was exchanged that day |
volume_btc | The volume exchanged in the cryptocurrency value that day, in BTC |
volume_currency | The volume exchanged in its converted value for that day, quoted in the corresponding fiat currency |
currency_code | Corresponds to the fiat currency used for non-BTC prices/volumes |
Finally, the currency_info
table maps the currency’s code to its English-language name:
Field | Description |
---|---|
currency_code | 2-7 character abbreviation for currency. Used in other hypertables |
currency | English name of currency |
When you have established the schema for the tables in the database, you can formulate create_table
SQL statements to actually create the tables you need:
--Schema for cryptocurrency analysis
DROP TABLE IF EXISTS "currency_info";
CREATE TABLE "currency_info"(
currency_code VARCHAR (10),
currency TEXT
);
--Schema for btc_prices table
DROP TABLE IF EXISTS "btc_prices";
CREATE TABLE "btc_prices"(
time TIMESTAMP WITH TIME ZONE NOT NULL,
opening_price DOUBLE PRECISION,
highest_price DOUBLE PRECISION,
lowest_price DOUBLE PRECISION,
closing_price DOUBLE PRECISION,
volume_btc DOUBLE PRECISION,
volume_currency DOUBLE PRECISION,
currency_code VARCHAR (10)
);
--Schema for crypto_prices table
DROP TABLE IF EXISTS "crypto_prices";
CREATE TABLE "crypto_prices"(
time TIMESTAMP WITH TIME ZONE NOT NULL,
opening_price DOUBLE PRECISION,
highest_price DOUBLE PRECISION,
lowest_price DOUBLE PRECISION,
closing_price DOUBLE PRECISION,
volume_crypto DOUBLE PRECISION,
volume_btc DOUBLE PRECISION,
currency_code VARCHAR (10)
);
--Schema for eth_prices table
DROP TABLE IF EXISTS "eth_prices";
CREATE TABLE "eth_prices"(
time TIMESTAMP WITH TIME ZONE NOT NULL,
opening_price DOUBLE PRECISION,
highest_price DOUBLE PRECISION,
lowest_price DOUBLE PRECISION,
closing_price DOUBLE PRECISION,
volume_eth DOUBLE PRECISION,
volume_currency DOUBLE PRECISION,
currency_code VARCHAR (10)
);
--Timescale specific statements to create hypertables for better performance
SELECT create_hypertable('btc_prices', 'time');
SELECT create_hypertable('eth_prices', 'time');
SELECT create_hypertable('crypto_prices', 'time');
Note that there are three create_hypertable
statements which are TimescaleDB-specific statements. A hypertable is an abstraction of a single continuous table across time intervals, so that you can query it using standard SQL. For more on hypertables, see the Timescale docs and this blog post.
Create a dataset to analyze
Now that you’ve defined the data you want, you can construct a dataset containing that data. You can write a small Python script for extracting data from CryptoCompare into four CSV files, called coin_names.csv
, crypto_prices.csv
, btc_prices.csv
, and eth_prices.csv
.
To get data from CryptoCompare, you’ll need to obtain an API key. For this analysis, the free key is sufficient.
The script consists of five parts:
- Import the necessary Python libraries in order to complete the data extraction
- Populate the
currency_info
table with a list of coin names - Get the historical Bitcoin (BTC) prices in 4198 other cryptocurrencies and populate the
crypto_prices
table - Get historical Bitcoin prices in different fiat currencies to populate
btc_prices
- Get historical Ethereum prices in different fiat currencies to populate
eth_prices
Here’s the full Python script, which you can also
#####################################################################
#1. Import library and setup API key
#####################################################################
import requests
import json
import csv
from datetime import datetime
apikey = 'YOUR_CRYPTO_COMPARE_API_KEY'
#attach to end of URLstring
url_api_part = '&api_key=' + apikey
#####################################################################
#2. Populate list of all coin names
#####################################################################
#URL to get a list of coins from cryptocompare API
URLcoinslist = 'https://min-api.cryptocompare.com/data/all/coinlist'
#Get list of cryptos with their symbols
res1 = requests.get(URLcoinslist)
res1_json = res1.json()
data1 = res1_json['Data']
symbol_array = []
cryptoDict = dict(data1)
#write to CSV
with open('coin_names.csv', mode = 'w') as test_file:
test_file_writer = csv.writer(test_file,
delimiter = ',',
quotechar = '"',
quoting=csv.QUOTE_MINIMAL)
for coin in cryptoDict.values():
if day.get('time') == None:
continue # skip this item
name = coin['Name']
symbol = coin['Symbol']
symbol_array.append(symbol)
coin_name = coin['CoinName']
full_name = coin['FullName']
entry = [symbol, coin_name]
test_file_writer.writerow(entry)
print('Done getting crypto names and symbols. See coin_names.csv for result')
#####################################################################
#3. Populate historical price for each crypto in BTC
#####################################################################
#Note: this part might take a while to run since we're populating data for 4k+ coins
#counter variable for progress made
progress = 0
num_cryptos = str(len(symbol_array))
for symbol in symbol_array:
# get data for that currency
URL = 'https://min-api.cryptocompare.com/data/histoday?fsym=' +
symbol +
'&tsym=BTC&allData=true' +
url_api_part
res = requests.get(URL)
res_json = res.json()
data = res_json['Data']
# write required fields into csv
with open('crypto_prices.csv', mode = 'a') as test_file:
test_file_writer = csv.writer(test_file,
delimiter = ',',
quotechar = '"',
quoting=csv.QUOTE_MINIMAL)
for day in data:
rawts = day['time']
ts = datetime.utcfromtimestamp(rawts).strftime('%Y-%m-%d %H:%M:%S')
o = day['open']
h = day['high']
l = day['low']
c = day['close']
vfrom = day['volumefrom']
vto = day['volumeto']
entry = [ts, o, h, l, c, vfrom, vto, symbol]
test_file_writer.writerow(entry)
progress = progress + 1
print('Processed ' + str(symbol))
print(str(progress) + ' currencies out of ' + num_cryptos + ' written to csv')
print('Done getting price data for all coins. See crypto_prices.csv for result')
#####################################################################
#4. Populate BTC prices in different fiat currencies
#####################################################################
# List of fiat currencies we want to query
# You can expand this list, but CryptoCompare does not have
# a comprehensive fiat list on their site
fiatList = ['AUD', 'CAD', 'CNY', 'EUR', 'GBP', 'GOLD', 'HKD',
'ILS', 'INR', 'JPY', 'KRW', 'PLN', 'RUB', 'SGD', 'UAH', 'USD', 'ZAR']
#counter variable for progress made
progress2 = 0
for fiat in fiatList:
# get data for bitcoin price in that fiat
URL = 'https://min-api.cryptocompare.com/data/histoday?fsym=BTC&tsym=' +
fiat +
'&allData=true' +
url_api_part
res = requests.get(URL)
res_json = res.json()
data = res_json['Data']
# write required fields into csv
with open('btc_prices.csv', mode = 'a') as test_file:
test_file_writer = csv.writer(test_file,
delimiter = ',',
quotechar = '"',
quoting=csv.QUOTE_MINIMAL)
for day in data:
rawts = day['time']
ts = datetime.utcfromtimestamp(rawts).strftime('%Y-%m-%d %H:%M:%S')
o = day['open']
h = day['high']
l = day['low']
c = day['close']
vfrom = day['volumefrom']
vto = day['volumeto']
entry = [ts, o, h, l, c, vfrom, vto, fiat]
test_file_writer.writerow(entry)
progress2 = progress2 + 1
print('processed ' + str(fiat))
print(str(progress2) + ' currencies out of 17 written')
print('Done getting price data for btc. See btc_prices.csv for result')
#####################################################################
#5. Populate ETH prices in different fiat currencies
#####################################################################
#counter variable for progress made
progress3 = 0
for fiat in fiatList:
# get data for bitcoin price in that fiat
URL = 'https://min-api.cryptocompare.com/data/histoday?fsym=ETH&tsym=' +
fiat +
'&allData=true' +
url_api_part
res = requests.get(URL)
res_json = res.json()
data = res_json['Data']
# write required fields into csv
with open('eth_prices.csv', mode = 'a') as test_file:
test_file_writer = csv.writer(test_file,
delimiter = ',',
quotechar = '"',
quoting=csv.QUOTE_MINIMAL)
for day in data:
rawts = day['time']
ts = datetime.utcfromtimestamp(rawts).strftime('%Y-%m-%d %H:%M:%S')
o = day['open']
h = day['high']
l = day['low']
c = day['close']
vfrom = day['volumefrom']
vto = day['volumeto']
entry = [ts, o, h, l, c, vfrom, vto, fiat]
test_file_writer.writerow(entry)
progress3 = progress3 + 1
print('processed ' + str(fiat))
print(str(progress3) + ' currencies out of 17 written')
print('Done getting price data for eth. See eth_prices.csv for result')
After running the script, you should have four .csv files:
python crypto_data_extraction.py
Load the dataset into TimescaleDB
Before you start, you need a working installation of TimescaleDB.
Set up the schema
Now all your hard work at the beginning comes in handy, and you can use the SQL script you created to set up the TimescaleDB unstance. If you don’t want to enter the SQL script by yourself, you can download
instead.
Log in to the TimescaleDB instance. Locate your host
, port
, and password
and then connect to the database:
psql -x "postgres://tsdbadmin:{YOUR_PASSWORD_HERE}@{YOUR_HOSTNAME_HERE}:{YOUR_PORT_HERE}/defaultdb?sslmode=require"
From the psql
command line, create a database. Let’s call it crypto_data
:
CREATE DATABASE crypto_data;
\c crypto_data
CREATE EXTENSION IF NOT EXISTS timescaledb CASCADE;
From the command prompt, you can apply the schema creation script to the database like this:
psql -x "postgres://tsdbadmin:{YOUR_PASSWORD_HERE}@{|YOUR_HOSTNAME_HERE}:{YOUR_PORT_HERE}/crypto_data?sslmode=require" < schema.sql
The output should look something like this:
NOTICE: 00000: table "currency_info" does not exist, skipping
LOCATION: DropErrorMsgNonExistent, tablecmds.c:1057
DROP TABLE
Time: 78.384 ms
CREATE TABLE
Time: 87.011 ms
NOTICE: 00000: table "btc_prices" does not exist, skipping
LOCATION: DropErrorMsgNonExistent, tablecmds.c:1057
DROP TABLE
Time: 77.094 ms
CREATE TABLE
Time: 79.815 ms
NOTICE: 00000: table "crypto_prices" does not exist, skipping
LOCATION: DropErrorMsgNonExistent, tablecmds.c:1057
DROP TABLE
Time: 78.430 ms
CREATE TABLE
Time: 78.430 ms
NOTICE: 00000: table "eth_prices" does not exist, skipping
LOCATION: DropErrorMsgNonExistent, tablecmds.c:1057
DROP TABLE
Time: 77.410 ms
CREATE TABLE
Time: 80.883 ms
create_hypertable
-------------------------
(1,public,btc_prices,t)
(1 row)
Time: 83.154 ms
create_hypertable
-------------------------
(2,public,eth_prices,t)
(1 row)
Time: 84.650 ms
create_hypertable
----------------------------
(3,public,crypto_prices,t)
(1 row)
Time: 81.864 ms
Now when you log back in to the TimescaleDB instance using psql
, you can run the \dt
command and see that the tables have been created properly:
List of relations
Schema | Name | Type | Owner
--------+---------------+-------+-----------
public | btc_prices | table | tsdbadmin
public | crypto_prices | table | tsdbadmin
public | currency_info | table | tsdbadmin
public | eth_prices | table | tsdbadmin
(4 rows)
Ingest data
Now that you’ve created the tables with the desired schema, all that’s left is to insert the data from the .csv files you created into the tables.
Make sure you are logged into TimescaleDB using psql
so that you can run each of these commands in turn:
\COPY btc_prices FROM btc_prices.csv CSV;
\COPY eth_prices FROM eth_prices.csv CSV;
\COPY crypto_prices FROM crypto_prices.csv CSV;
\COPY currency_info FROM coin_names.csv CSV;
important
Data ingestion could take a while, depending on the speed of your Internet connection.
You can verify that the ingestion worked by running a simple SQL command, such as:
SELECT * FROM btc_prices LIMIT 5;
You should get something like this output:
-[ RECORD 1 ]---+-----------------------
time | 2013-03-11 00:00:00+00
opening_price | 60.56
highest_price | 60.56
lowest_price | 60.56
closing_price | 60.56
volume_btc | 0.1981
volume_currency | 12
currency_code | AUD
-[ RECORD 2 ]---+-----------------------
time | 2013-03-12 00:00:00+00
opening_price | 60.56
highest_price | 60.56
lowest_price | 41.38
closing_price | 47.78
volume_btc | 47.11
volume_currency | 2297.5
currency_code | AUD
-[ RECORD 3 ]---+-----------------------
time | 2013-03-07 00:00:00+00
opening_price | 181.15
highest_price | 273.5
lowest_price | 237.4
closing_price | 262.87
volume_btc | 33.04
volume_currency | 8974.45
currency_code | CNY
-[ RECORD 4 ]---+-----------------------
time | 2013-03-07 00:00:00+00
opening_price | 32.31
highest_price | 35.03
lowest_price | 26
closing_price | 31.57
volume_btc | 13321.61
volume_currency | 425824.38
currency_code | EUR
-[ RECORD 5 ]---+-----------------------
time | 2013-03-11 00:00:00+00
opening_price | 35.7
highest_price | 37.35
lowest_price | 35.4
closing_price | 37.15
volume_btc | 3316.09
volume_currency | 121750.98
currency_code | EUR
Time: 224.741 ms
Query and analyze the data
At the beginning of the tutorial, we defined some questions to answer. Naturally, each of those questions has an answer in the form of a SQL query. Now that you database is set up properly, and the data is captured and ingested, you can find some answers:
For example, How did Bitcoin price in USD vary over time?
SELECT time_bucket('7 days', time) AS period,
last(closing_price, time) AS last_closing_price
FROM btc_prices
WHERE currency_code = 'USD'
GROUP BY period
ORDER BY period
How did BTC daily returns vary over time? Which days had the worst and best returns?
SELECT time,
closing_price / lead(closing_price) over prices AS daily_factor
FROM (
SELECT time,
closing_price
FROM btc_prices
WHERE currency_code = 'USD'
GROUP BY 1,2
) sub window prices AS (ORDER BY time DESC)
How did the trading volume of Bitcoin vary over time in different fiat currencies?
SELECT time_bucket('7 days', time) AS period,
currency_code,
sum(volume_btc)
FROM btc_prices
GROUP BY currency_code, period
ORDER BY period
How did Ethereum (ETH) price in BTC vary over time?
SELECT
time_bucket('7 days', time) AS time_period,
last(closing_price, time) AS closing_price_btc
FROM crypto_prices
WHERE currency_code='ETH'
GROUP BY time_period
ORDER BY time_period
How did ETH prices, in different fiat currencies, vary over time?
SELECT time_bucket('7 days', c.time) AS time_period,
last(c.closing_price, c.time) AS last_closing_price_in_btc,
last(c.closing_price, c.time) * last(b.closing_price, c.time) FILTER (WHERE b.currency_code = 'USD') AS last_closing_price_in_usd,
last(c.closing_price, c.time) * last(b.closing_price, c.time) FILTER (WHERE b.currency_code = 'EUR') AS last_closing_price_in_eur,
last(c.closing_price, c.time) * last(b.closing_price, c.time) FILTER (WHERE b.currency_code = 'CNY') AS last_closing_price_in_cny,
last(c.closing_price, c.time) * last(b.closing_price, c.time) FILTER (WHERE b.currency_code = 'JPY') AS last_closing_price_in_jpy,
last(c.closing_price, c.time) * last(b.closing_price, c.time) FILTER (WHERE b.currency_code = 'KRW') AS last_closing_price_in_krw
FROM crypto_prices c
JOIN btc_prices b
ON time_bucket('1 day', c.time) = time_bucket('1 day', b.time)
WHERE c.currency_code = 'ETH'
GROUP BY time_period
ORDER BY time_period
Which cryptocurrencies had the most transaction volume in the past 14 days?
SELECT 'BTC' AS currency_code,
sum(b.volume_currency) AS total_volume_in_usd
FROM btc_prices b
WHERE b.currency_code = 'USD'
AND now() - date(b.time) < INTERVAL '14 day'
GROUP BY b.currency_code
UNION
SELECT c.currency_code AS currency_code,
sum(c.volume_btc) * avg(b.closing_price) AS total_volume_in_usd
FROM crypto_prices c JOIN btc_prices b ON date(c.time) = date(b.time)
WHERE c.volume_btc > 0
AND b.currency_code = 'USD'
AND now() - date(b.time) < INTERVAL '14 day'
AND now() - date(c.time) < INTERVAL '14 day'
GROUP BY c.currency_code
ORDER BY total_volume_in_usd DESC
Which cryptocurrencies had the top daily return?
WITH
prev_day_closing AS (
SELECT
currency_code,
time,
closing_price,
LEAD(closing_price) OVER (PARTITION BY currency_code ORDER BY TIME DESC) AS prev_day_closing_price
FROM
crypto_prices
)
, daily_factor AS (
SELECT
currency_code,
time,
CASE WHEN prev_day_closing_price = 0 THEN 0 ELSE closing_price/prev_day_closing_price END AS daily_factor
FROM
prev_day_closing
)
SELECT
time,
LAST(currency_code, daily_factor) AS currency_code,
MAX(daily_factor) AS max_daily_factor
FROM
daily_factor
GROUP BY
time
Next steps
While it’s fun to run SQL queries in the command line, the real magic is when you’re able to visualize it. Follow the companion tutorial to this piece and learn how to use TimescaleDB and Tableau together to visualize your time-series data.
Ready for even more learning? Here’s a few suggestions: