Today, I'll take you to learn about a minority , But a powerful visualization Library mplfinance, Master the most flexible... Together python Library to create a beautiful financial Visualization .
The proliferation of programming and technology applications in the financial sector is inevitable , Growth never seems to decline . One of the most interesting parts of application programming is the interpretation and visualization of historical or real-time stock data .
Now? , In order to be in python Visualize general data in ,matplotlib、seaborn When the module starts to work , however , When it comes to visualizing financial data ,Plotly Will be the first choice , Because it provides built-in functions with interactive visual effects . Here I want to introduce an unknown hero , It's just mplfinance library matplotlib Brother Library .
We all know matplotlib The versatility of the package , And you can easily draw any type of data . Even financial charts like candlesticks can be used matplotlib Package drawing , But we have to start from scratch .
lately , I began to know that there was a man named mplfinance Separate modules , Designed to create advanced financial visualization . In this paper , We will delve into this Python library , And explore its function of generating different types of charts .
Import the required packages into our python Environment is an essential step . In this paper , We need three bags , They process data frames Pandas、 call API And extracting stock data requests, And creating financial charts mplfinance. For those who have not yet installed these packages , Please copy this code to your terminal ：
pip install pandas pip install requests pip install mplfinance
After installing the package , It's time to import them into our python In the environment .
import pandas as pd import requests import mplfinance as mf
Now? , We have imported all the necessary packages . Let's use 12data.com[1] Provided API The endpoint pulls Amazon's historical stock data . Before that ,12data.com A note on ：12data Is one of the leading market data providers , Have a large amount of data for all types of markets API Endpoint . With twelve data provided API Interaction is very easy , And one of the best documents ever . Besides , Please make sure you are in 12data.com Have an account on , That's the only way , You can access your API secret key （ Use API Extract important elements of data ）.
def get_historical_data(symbol, start_date):
api_key = 'YOUR API KEY'
api_url = f'https://api.twelvedata.com/time_series?symbol={symbol}&interval=1day&outputsize=5000&apikey={api_key}'
raw_df = requests. get(api_url).json()
df = pd.DataFrame(raw_df['values']).iloc[::-1].set_index('datetime').astype(float)
df = df[df.index >= start_date]
df.index = pd.to_datetime(df.index)
return df
amzn = get_historical_data('AMZN', '2021-01-01')
amzn.tail() The first thing we do is define a name called 'get_historical_data' Function of , This function takes the stock code （'symbol'） And the start date of historical data （'start_date'） Is the parameter .
Inside the function , We defined API Key and URL, And store them in their own variables .
Next , We use 'get' Function to JSON Format extracts historical data and stores it in 'raw_df' variable . On the original JSON After some data cleaning and formatting , We use an empty Pandas DataFrame Return it as .
Last , We call created Function to pull Amazon from 2021 Historical data from the beginning of the year , And store it to "amzn" variable .
OHLC A chart is a bar chart , Displays the opening price for each period 、 Highest price 、 The lowest price and closing price .
OHLC Charts are useful , Because they show four main data points over a period of time , Many traders think the closing price is the most important . It also helps to show increased or decreased momentum . Strong performance when opening and closing are far apart , When the opening and closing are similar, it shows indecision or weak kinetic energy .
The highest and lowest prices show the full price range for the period , Helps to assess volatility 1[2]. Now I'm going to use mplfinance Create a OHLC Chart , Just one line of code ：
mf.plot(amzn.iloc[:-50,:])
In the code above , We first call the plot function , And in which we extracted Amazon OHLC The data slice is the last 50 Readings , The purpose of this is just to make the chart clearer , So that the element is visible . The single line code above will produce the output shown below ：
Traders use candlestick charts to determine possible price changes based on past patterns . Candlesticks are very useful in trading , Because they show four price points throughout the time period specified by the trader （ Opening price 、 Closing price 、 The highest price and the lowest price ）.
The most interesting part of this type of chart is that it can also help traders read emotions , This is the primary driver of the market itself 2[3]. To use mplfinance Generate Candlestick diagram , We just need to add another parameter , That is, the of the function type Parameters plot and candle Mention in it . The code is as follows ：
mf.plot(amzn.iloc[:-50,:], type = 'candle')
The above code will generate a candlestick chart as shown below ：
Brick diagram ( Renko chart) Is a chart constructed using price changes , Instead of using both prices and standardized intervals like most charts . The chart looks like a series of bricks , When the price moves the specified price amount, a new brick is created , And each block is made of the previous brick 45 Degree angle （ Up or down ）.Renko The main purpose of charts is to filter out noise and help traders see trends more clearly , Because all motions smaller than the frame size are filtered out 3[4] .
as far as I am concerned ,mplfinance Is the only provider Renko Chart's Python library , That's what we're going to see next , That's why this package has a strong advantage in financial Visualization . Now create a Renko, We just need to be in the function renko Of type Specified in parameter plot.Renko The code of the chart is as follows ：
mf.plot(amzn, type = 'renko')
We can also ask plot Add an extra argument to the function , This parameter is based on renko_params We need and other similar types to modify the parameters of brick size , But I prefer the default . The above code generates a brick diagram that looks like this ：
Dot graph , abbreviation P&F chart , Be similar to Renko chart , It plots the price trend of assets without considering the passage of time . With some other types of charts （ For example, Candlestick ） contrary , The candlestick indicates the degree of change of assets within a set period of time , and P&F The chart uses a stack of X or O The columns that make up , Each column represents a certain number of price changes .X Represents a rise in prices , and O Represents a decline in prices . When the price reverses, the reverse volume 4[5] when , Will be in O Then form a new X Column or in X Then form a new O Column .
The function supporting point graph cannot be found elsewhere , Only in mplfinance Found in Library , And it also allows us to pass only pnf In function type The process of creating a chart specified in the parameter is easier plot. The code is as follows ：
mf.plot(amzn, type = 'pnf')
Point chart add more information
mplfinance Packages are not limited to generating different types of charts , It also allows us to add a simple moving average (SMA) Additional indicators such as and trading volume make these charts more insightful . For those who don't know both , Trading volume is the number of stocks bought and sold by traders within a specific time frame , And the simple moving average (SMA) It's just the average price over a specific period of time . It is a technical index , Widely used to create trading strategies .
use matplotlib It takes a thousand years to plot these data , and mplfinance Allow us to do this in one line of code . except type Besides the parameters , We just need to introduce two other parameters , One is mav We must specify each SMA Parameters of the backtracking period , The other is volume The parameters we must mention ,True If we want to add the volume chart to our chart , perhaps False We don't want to . The codes of these two indicators are as follows ：
mf.plot(amzn, mav = (10, 20), type = 'candle', volume = True)
The above code can be modified and tested in two ways . The first approach is obviously to try different types of charts . In the above code , We mentioned that our chart type is Candlestick , But you can change it to OHLC、Renko even to the extent that P&F Chart , And observe the appearance of each chart and its two additional indicators . The next method is to use mav We can add any number of... With different review periods SMA Parameters of . The output of the above code is shown below ：
If you want to know how to save any of these financial visualizations , Just add another parameter ,savefig That is, you only need to mention the parameters of its file name , The rest will be processed . Suppose you want to save the picture above , Then the code you must follow is as follows ：
mf.plot(amzn, mav = (10, 20), type = 'candle', volume = True, savefig = 'amzn.png')
That's all you need to do to save a wonderful financial Visualization . be prone to , Right ？
in my opinion , And Plotly or Altair Wait for the library to compare ,mplfinance It is the most powerful library for drawing financial data . This article simply introduces the use of mplfinance Functions that can be realized , But this great library comes with many new features . It allows us to add custom technical indicator data , And draw with the actual chart , We can customize the whole template , Even every element in the chart , Add trendline , wait .
The best part of this library is its ease of use , And help us generate high-level financial visualization with one line of code . Though like Plotly Such packages have built-in functions to create these charts , But it can't be done in one line of code .
mplfinance The only drawback now is its poor documentation , This makes people don't even know what the bag is about . Documentation is a crucial aspect , When it comes to open source projects , Documentation should be considered critical . Special like mplfinance Such critical and useful projects must be clearly documented , Have a clear explanation of the tools and functions it provides .
Come here , After you read this article . If you forget the code of the chart , Don't worry about , Finally, I provide the complete source code . You can also collect this article , Check it when you need it .
import pandas as pd
import requests
import mplfinance as mf
# Extracting stock data
def get_historical_data(symbol, start_date):
api_key = 'YOUR API KEY'
api_url = f'https://api.twelvedata.com/time_series?symbol={symbol}&interval=1day&outputsize=5000&apikey={api_key}'
raw_df = requests.get(api_url).json()
df = pd.DataFrame(raw_df['values']).iloc[::-1].set_index('datetime').astype(float)
df = df[df.index >= start_date]
df.index = pd.to_datetime(df.index)
return df
amzn = get_historical_data('AMZN', '2021-01-01')
amzn.tail()
# 1. OHLC Chart
mf.plot(amzn.iloc[:-50,:])
# 2. Candlestick Chart
mf.plot(amzn.iloc[:-50,:], type = 'candle')
# 3. Renko Chart
mf.plot(amzn, type = 'renko')
# 4. Point and Figure Chart
mf.plot(amzn, type = 'pnf')
# 5. Technical chart
mf.plot(amzn, mav = (10, 20), type = 'candle', volume = True)
# 6. Plot customization
mf.plot(amzn, mav = (5, 10, 20), type = 'candle',
volume = True, figratio = (10,5),
style = 'binance', title = 'AMZN STOCK PRICE',
tight_layout = True)
# 7. Saving the plot
mf.plot(amzn, mav = (5, 10, 20), type = 'candle',
volume = True, figratio = (10,5),
style = 'binance', title = 'AMZN STOCK PRICE',
tight_layout = True, savefig = 'amzn.png')
