Building a DEX Tracker with Streamlit and Python

In this tutorial, we'll explore how to connect to CoinGecko API to query on-chain, decentralized exchange (DEX) data to create a DEX tracker using Streamlit and Python.

How to Build a DEX Tracker in 5 Steps

Step 1. Setup

To get data on all the newest pools across all supported blockchain networks, we’ll query the New Pools List endpoint. Such data can be particularly useful for crypto traders to identify new tokens and projects entering the market, evaluate the risk associated with each liquidity pool and spot potential arbitrage opportunities.

For this project, I would suggest using a virtual environment to install and isolate your dependencies.

Step 2. Boilerplate

We used this to get started: https://github.com/streamlit/app-starter-kit

If you want to just copy the code directly, you can download this repo on GitHub for the quickest setup. 

Step 3. Display + API Utilities

To fetch and parse the API, you’ll first need to get the associated data, followed by parsing out the raw JSON response into a format acceptable for streamlit. In this example, we created a separate utils.py file to handle these various functions.

fetch_data is a generic decorated function for APIs.

process_data takes our newly captured API JSON response and converts it into a data frame.

get_top_changes specifically gets the top3 pairs with the greatest change in a specific column. In this example, we use price_change_percentage_h1

parse_crypto_pools is used to map the response to a flat format. 

plot_price_change to plot the price changes in the last 24 hours

Here's how it all comes together:

import json

import streamlit as st

import requests

import pandas as pd

import matplotlib.pyplot as plt

import seaborn as sns 

# Function to fetch data from the API

@st.cache_data

def fetch_data(url, headers):

   response = requests.get(url, headers=headers)

   if response.status_code == 200:

       return response.json()

   else:

       st.error(f'Failed to retrieve data. Status code: {response.status_code}')

       return None

# Function to process the JSON data into a DataFrame

def process_data(data):

   return pd.DataFrame([pool['attributes'] for pool in data['data']])

# Function to get top 3 pairs with the greatest 1-hour volume change

def get_top_changes(df, column_name='price_change_percentage_h1'):

   # Ensure the column is of type float

   df[column_name] = df[column_name].astype(float)

   # Sort the DataFrame based on the volume change column

   top_changes = df.sort_values(by=column_name, ascending=False).head(3)

   # Include the 'transactions' column in the returned DataFrame

   return top_changes[['name', column_name, 'transactions_h1_sells', 'transactions_h1_buys']]

def parse_crypto_pools(json_payload):

   # Convert the JSON payload to a Python dictionary

   data = json_payload

   # List to hold parsed pool data

   parsed_pools = []

   # Iterate through each pool in the data

   for pool in data['data']:

       # Extract relevant information from each pool

       pool_info = {

           'address': pool['attributes']['address'],

           'name': pool['attributes']['name'],

           'pool_created_at': pool['attributes']['pool_created_at'],

           'fdv_usd': pool['attributes'].get('fdv_usd'),

           'market_cap_usd': pool['attributes'].get('market_cap_usd'),

           'volume_usd_h1': pool['attributes']['volume_usd']['h1'],

           'volume_usd_h24': pool['attributes']['volume_usd']['h24'],

           'price_change_percentage_h1': json.loads(pool['attributes']['price_change_percentage']['h1']),

           'price_change_percentage_h24': json.loads(pool['attributes']['price_change_percentage']['h24']),

           'transactions_h1_buys': pool['attributes']['transactions']['h1']['buys'],

           'transactions_h1_sells': pool['attributes']['transactions']['h1']['sells'],

       }

       parsed_pools.append(pool_info)

   return parsed_pools

def plot_price_change(df, log_scale=False):

   pool_names = df['name']

   # Directly use the float values from the series

   price_changes_24h = [float(change) if change else 0 for change in df['price_change_percentage_h1']]

   fig, ax = plt.subplots()

   sns.barplot(x='price_change_percentage_h1', y='name', data=df, ax=ax)

   ax.set(xlabel='Price Change Percentage (24h)', title='Price Change Percentage (24h) by Pool')

   return fig

4. Main Dashboard

For the main.py, we’ll reserve this for logic related to constants, UI configuration, and running the main application. 

import streamlit as st

import pandas as pd

import numpy as np

from utils import fetch_data, get_top_changes, parse_crypto_pools, plot_price_change

# Constants

BASE_URL = 'https://pro-api.coingecko.com/api/v3/onchain'

ENDPOINT = '/networks/new_pools?page=1'

API_URL = BASE_URL + ENDPOINT

HEADERS = {'accept': 'application/json', 'x-cg-pro-api-key':'INSERT_API_KEY_HERE'}

# Set the page to wide mode

st.set_page_config(layout="wide")

# Main app code

st.title('Cryptocurrency Pool Data Viewer')

# Fetch and process data

raw_data = fetch_data(API_URL, HEADERS)

if raw_data:

   df = pd.DataFrame(parse_crypto_pools(raw_data))

   top_volume_changes = get_top_changes(df, 'price_change_percentage_h1')

   # Display the top 3 price changes

   st.subheader('⭐️ Top 3 Pairs by 1-Hour Price Change')

   columns = st.columns(3)  # Create three columns for the top 3 changes

   for index, (col, row) in enumerate(zip(columns, top_volume_changes.iterrows())):

       with col:

           st.metric(label=row[1]['name'], value=f"{row[1]['price_change_percentage_h1']:.2f}%")

           st.text(f"Buys in last hour: {row[1]['transactions_h1_buys']}")

           st.text(f"Sells in last hour: {row[1]['transactions_h1_sells']}")

   # Create two columns for the data table and the bar chart

   col1, col2 = st.columns([2,3])

   with col1:

       # Display data table in the first column

       st.subheader('🏓 Data Overview')

       st.write("", df)

   with col2:

       # Visualization: Bar Chart for Price Change Percentage in the second column

       st.subheader("📊 Price Change Percentage (24h)")

       if 'price_change_percentage_h1' in df.columns:

           fig_price_change = plot_price_change(df)

           st.pyplot(fig_price_change, use_container_width=True)

The Final Result: Crypto Pools DEX Tracker on Streamlit

If you’ve setup your dependencies and code properly, you can now run streamlit run main.py in your terminal, after which the below dashboard should appear in your browser:

By following these steps, you can easily build a DEX dashboard to visualize on-chain data from decentralized exchanges. This guide serves as a foundation, and you can expand it by adding more features and interactivity.

Enhancing Your Dashboard: 5 Popular On-Chain DEX API Endpoints

For more details on enhancing your dashboard, you may refer to this Streamlit documentation as well as leverage other popular on-chain DEX data endpoints on CoinGecko API, exclusively for paid subscribers:

1. /onchain/networks/{network}/pools – Get top pools on a network
2. /onchain/networks/{network}/trending_pools – Get trending pools on a network
3. /onchain/networks/{network}/tokens/{token_address}/pools – Get top pools for a token, with a provided token contract address
4. /onchain/search/pools – Search for pool address, token address or token symbols on a network
5. /onchain/networks/{network} /pools/{pool_address}/ohlcv/{timeframe} – Get the OHLCV chart (Open, High, Low, Close, Volume) of a crypto pool based on a pool address on a network.

Interested in more how-to guides leveraging CoinGecko API? Check out this walk-through on how to build an artificial intelligence crypto chatbot.

CoinGecko's Content Editorial Guidelines

CoinGecko’s content aims to demystify the crypto industry. While certain posts you see may be sponsored, we strive to uphold the highest standards of editorial quality and integrity, and do not publish any content that has not been vetted by our editors.

Learn more

Automation Architech

Automation Architech boasts a seasoned team specializing in ML-driven solutions, operations engineering, and digital product development. They are dedicated to ethical AI, bridging tech gaps, and fostering creativity without sidelining humanity. Follow the author on Twitter @auto_architech