Drawing DEX prices using OHLCV candle charts#

This is an example notebook how to draw a price candle chart for Trading Strategy.
Please read how to set up your environment first.
We draw an OHLCV chart. Internally trading-strategy is using Plotly framework for charting.

Getting started#

First, let’s create Trading Strategy oracle market data client. If you do not have an API key yet, you will be asked to create one.

[8]:

from tradingstrategy.client import Client

client = Client.create_jupyter_client()

Started Trading Strategy in Jupyter notebook environment, configuration is stored in /Users/moo/.tradingstrategy

Resolve trading pair metadata#

We look at BNB / BUSD trading pair on PancakeSwap DEX (version 2).

To download the data we first need to download exchange and trading pair metadata, so that we can resolve symbolic exchange name and ticket information to the trading pair id
We may have duplicate trading pair with the same ticket and token symbol information. This is often the case for popular trading pairs as there are a lot of scam pairs around with the same symbol. We go around this by using the option pick_by_highest_vol which gives us the correct trading pair.
Note that the blockchain native token of EVM chain is presented by its wrapped version on DEXes. E.g. BNB on BNB Chain becomes WBNB.

[9]:

from pyarrow import Table
from tradingstrategy.chain import ChainId
from tradingstrategy.exchange import ExchangeUniverse
from tradingstrategy.pair import PandasPairUniverse, DEXPair

# Fetch all exchange names, slugs and addresses
exchange_universe: ExchangeUniverse = client.fetch_exchange_universe()

# Fetch all trading pairs across all exchanges
pair_table: Table = client.fetch_pair_universe()
pair_universe = PandasPairUniverse(pair_table.to_pandas())

pancake_v2 = exchange_universe.get_by_chain_and_slug(ChainId.bsc, "pancakeswap-v2")

pair: DEXPair = pair_universe.get_one_pair_from_pandas_universe(
    pancake_v2.exchange_id,
    "WBNB",
    "BUSD",
    pick_by_highest_vol=True)

print(f"PancakeSwap v2 is {pancake_v2}")
print(f"We have trading pair {pair}")

PancakeSwap v2 is <Exchange PancakeSwap v2 at 0xca143ce32fe78f1f7019d7d551a6402fc5350c73 on Binance Smart Chain>
We have trading pair <Pair #1364760 WBNB - BUSD (0x58f876857a02d6762e0101bb5c46a8c1ed44dc16) at exchange #1187 on binance>

Fetch price data feed#

We then downlaod OHLCV data using Trading Strategy oracle’s real-time API endpoint
The data is internally delivered as JSONL, but this is abstracted way by trading-strategy Python client library
Any price of non-dollar pairs, e.g. AAVE/ETH, is pre-converted to US dollars by Trading Strategy exchange rate rules

[10]:

import pandas as pd
from tradingstrategy.timebucket import TimeBucket

candles: pd.DataFrame = client.fetch_candles_by_pair_ids(
    {pair.pair_id},
    TimeBucket.d1,
    progress_bar_description=f"Download data for {pair.get_ticker()}"
)

Inspect OHCLV data#

You can manually inspect OHCLV trading dataframes.
Easiest way to do this is to use IPython display() function to render dataframes as table.

[11]:

# Display latest 14 days of WBNB-BUSD data
display(candles.iloc[-14:])

	pair_id	timestamp	open	high	low	close	buy_volume	sell_volume	start_block	end_block	exchange_rate	buys	sells	avg	volume
timestamp
2022-11-23	1364760	2022-11-23	266.261026	299.841244	264.315010	297.825673	2.729343e+07	2.228333e+07	23288796	23317233	1.0	21300.0	24550.0	NaN	4.957676e+07
2022-11-24	1364760	2022-11-24	297.822293	302.814980	292.549276	300.091218	9.670815e+06	9.369412e+06	23317237	23345600	1.0	16895.0	16145.0	NaN	1.904023e+07
2022-11-25	1364760	2022-11-25	300.087669	304.630323	293.529879	300.225753	9.101830e+06	9.063683e+06	23345606	23373804	1.0	16136.0	14800.0	NaN	1.816551e+07
2022-11-26	1364760	2022-11-26	300.225295	316.787367	300.139182	311.051139	1.072095e+07	9.386321e+06	23373806	23402018	1.0	15987.0	15941.0	NaN	2.010727e+07
2022-11-27	1364760	2022-11-27	311.050841	316.365007	306.025563	306.773434	9.036277e+06	9.531502e+06	23402024	23430502	1.0	15283.0	14602.0	NaN	1.856778e+07
2022-11-28	1364760	2022-11-28	306.773488	308.490684	289.239530	293.141829	1.015980e+07	1.179148e+07	23430504	23458977	1.0	17488.0	14359.0	NaN	2.195128e+07
2022-11-29	1364760	2022-11-29	293.141677	306.658445	289.679437	295.772926	6.853360e+06	6.512749e+06	23458979	23487565	1.0	16409.0	15357.0	NaN	1.336611e+07
2022-11-30	1364760	2022-11-30	295.772956	305.356409	293.931451	300.922936	9.490703e+06	8.836190e+06	23487568	23516178	1.0	19505.0	17381.0	NaN	1.832689e+07
2022-12-01	1364760	2022-12-01	300.924183	301.409985	290.616758	291.678394	9.108918e+06	1.030238e+07	23516180	23544574	1.0	20378.0	17129.0	NaN	1.941129e+07
2022-12-02	1364760	2022-12-02	291.681450	293.328217	286.174944	292.979351	1.483278e+07	1.458857e+07	23544576	23572636	1.0	22339.0	18104.0	NaN	2.942134e+07
2022-12-03	1364760	2022-12-03	292.979907	293.566933	289.695106	290.307121	5.278060e+06	5.676123e+06	23572640	23600811	1.0	16711.0	14991.0	NaN	1.095418e+07
2022-12-04	1364760	2022-12-04	290.306878	294.650247	288.330343	292.470853	4.900642e+06	4.552601e+06	23600814	23628935	1.0	17466.0	13406.0	NaN	9.453243e+06
2022-12-05	1364760	2022-12-05	292.470940	297.145499	286.615769	287.910945	5.476807e+06	6.183128e+06	23628937	23657267	1.0	17560.0	13305.0	NaN	1.165994e+07
2022-12-06	1364760	2022-12-06	287.909748	292.217899	287.497567	289.707345	2.083389e+06	1.794518e+06	23657271	23667239	1.0	4735.0	3841.0	NaN	3.877908e+06

Draw candle chart#

Display the price chart using OHLCV candles
We have a shortcut function for this, but you can as well construct Plotly chart by hand

[12]:

from tradingstrategy.charting.candle_chart import visualise_ohlcv

figure = visualise_ohlcv(
    candles,
    chart_name=f"{pair.base_token_symbol} - {pair.quote_token_symbol} price chart",
    y_axis_name=f"$ {pair.base_token_symbol} price",
)

display(figure)

Dark theme for the price chart#

You can update Plotly figure settings to change its colour theme.
See Plotly default colour scheme and theming options.

[13]:

figure.update_layout(template="plotly_dark")

display(figure)

Unlabeled chart#

We can create a chart without any labels as well.

[14]:

figure = visualise_ohlcv(
    candles,
    chart_name=None,
    y_axis_name=None,
    volume_axis_name=None,
)

display(figure)