CCXT_Predict_Bitcoin_from_Binance.ipynb

Kernel: Python 3

CCXT - Predict Bitcoin from Binance

Give Feedback | Bug report

Tags: #ccxt #bitcoin #trading #investors #ai #analytics #plotly

Author: Jeremy Ravenel

Last update: 2023-04-12 (Created: 2021-02-28)

Description: This notebook uses the CCXT library to predict Bitcoin prices on the Binance exchange.

Input

Install package

!pip install ccxt --user

Import libraries

import naas
import ccxt
import pandas as pd
from datetime import datetime
from naas_drivers import plotly, prediction

Setup Binance

👉 How to create API ?

binance_api = ""
binance_secret = ""

Variables

symbol = "BTC/USDT"
limit = 200
timeframe = "1d"

Model

Get data

binance = ccxt.binance({"apiKey": binance_api, "secret": binance_secret})

data = binance.fetch_ohlcv(symbol=symbol, limit=limit, timeframe=timeframe)

Mapping of the candlestick plot

df = pd.DataFrame(data, columns=["Date", "Open", "High", "Low", "Close", "Volume"])
df["Date"] = [datetime.fromtimestamp(float(time) / 1000) for time in df["Date"]]

Attribute the candlesticks variables

chart_candlestick = plotly.candlestick(
    df,
    label_x="Date",
    label_open="Open",
    label_high="High",
    label_low="Low",
    label_close="Close",
)

df[f"MA{20}"] = df.Close.rolling(20).mean()
df[f"MA{50}"] = df.Close.rolling(50).mean()

Get the prediction for the stock plot

pr = prediction.get(dataset=df)
chart_stock = plotly.stock(pr, kind="linechart")

Output

Display bitcoin plot prediction by changing resolution

chart_stock.update_layout(
    autosize=True,
    width=1300,
    height=800,
)