Home / Questions / 556473
Accepted
NoahVerner
Asked: 2022-09-12 15:39:35 +0800 CST 2022-09-12 15:39:35 +0800 CST

How to properly fit grid lines on a Japanese candlestick chart using MatPlotLibFinance in Python 3?

  • 772

I am currently facing an issue related to displaying x-axis data on a candlestick chart.

In short words:

The plotted vertical grid lines do not line up with the candlesticks.

Some of these vertical grid lines are offset per 1candlestick after the first Japanese candlestick.

As can be seen in the graph below:

problem

The dataframe (called df_trading_pair_date_time_index) used to graph the above was the following:

                      Open   High    Low  Close   Volume                End Date
Start Date                                                                      
2022-09-11 11:45:00  7.696  7.704  7.695  7.704    267.3 2022-09-11 11:47:59.999
2022-09-11 11:48:00  7.703  7.703  7.695  7.697    537.5 2022-09-11 11:50:59.999
2022-09-11 11:51:00  7.696  7.701  7.695  7.695    292.7 2022-09-11 11:53:59.999
2022-09-11 11:54:00  7.695  7.699  7.695  7.699     64.8 2022-09-11 11:56:59.999
2022-09-11 11:57:00  7.701  7.702  7.701  7.702    101.5 2022-09-11 11:59:59.999
2022-09-11 12:00:00  7.702  7.705  7.680  7.685   4487.0 2022-09-11 12:02:59.999
2022-09-11 12:03:00  7.687  7.691  7.681  7.689   2482.5 2022-09-11 12:05:59.999
2022-09-11 12:06:00  7.691  7.691  7.677  7.682   1074.8 2022-09-11 12:08:59.999
2022-09-11 12:09:00  7.686  7.709  7.686  7.709   1389.5 2022-09-11 12:11:59.999
2022-09-11 12:12:00  7.708  7.708  7.690  7.701   1277.5 2022-09-11 12:14:59.999
2022-09-11 12:15:00  7.698  7.698  7.690  7.695   1655.7 2022-09-11 12:17:59.999
2022-09-11 12:18:00  7.694  7.698  7.692  7.696    576.2 2022-09-11 12:20:59.999
2022-09-11 12:21:00  7.696  7.700  7.694  7.696   1711.3 2022-09-11 12:23:59.999
2022-09-11 12:24:00  7.698  7.703  7.697  7.703   1893.1 2022-09-11 12:26:59.999
2022-09-11 12:27:00  7.701  7.720  7.701  7.717   1256.0 2022-09-11 12:29:59.999
2022-09-11 12:30:00  7.718  7.743  7.718  7.739   5602.6 2022-09-11 12:32:59.999
2022-09-11 12:33:00  7.739  7.744  7.730  7.744    905.6 2022-09-11 12:35:59.999
2022-09-11 12:36:00  7.743  7.780  7.740  7.780   6651.3 2022-09-11 12:38:59.999
2022-09-11 12:39:00  7.780  7.782  7.769  7.772   4478.3 2022-09-11 12:41:59.999
2022-09-11 12:42:00  7.773  7.886  7.773  7.875  30280.6 2022-09-11 12:44:59.999

The data types of the columns are as follows:

Start Date  datetime64[ns]
Open               float64
High               float64
Low                float64
Close              float64
Volume             float64
End Date    datetime64[ns]

The code I wrote to graph the above was as follows:

import pandas as pd
import mplfinance as mpf    

trading_pair= "DOTBUSD"

# Plotting
# Create my own `marketcolors` style:
mc = mpf.make_marketcolors(up='#2fc71e',down='#ed2f1a',inherit=True)
# Create my own `MatPlotFinance` style:
s  = mpf.make_mpf_style(base_mpl_style=['bmh', 'dark_background'],marketcolors=mc, y_on_right=True)    

# Plot it
trading_plot, axlist = mpf.plot(df_trading_pair_date_time_index,
                    figratio=(10, 6),
                    type="candle",
                    style=s,
                    tight_layout=True,
                    datetime_format = '%H:%M',
                    ylabel = "Precio ($)",
                    returnfig=True
                    )
# Add Title
symbol = trading_pair.replace("BUSD","")+"/"+"BUSD"
axlist[0].set_title(f"{symbol} - 3m", fontsize=25, style='italic', fontfamily='fantasy')

So I decided to open this question to learn how to properly adjust the grid lines for plotting candlestick data, after reading this other similar question I tried changing the values ​​of figratioa (12,6)and (20,7)but that didn't solve my problem at all.

I just wish my chart ended up looking something like this, not necessarily covering every candle, but making sure to align some of the x-axis data with their corresponding candles:

desired output

python-3.x

1 Answers

  1. Best Answer

    Good day,

    The easiest way I could think of to adapt the solution to your code is by using the argument show_nontrading = Truetogether with matplotlib.axis.Axis.set_major_locatorandmatplotlib.axis.Axis.set_minor_locator

    Full example using your data from the "sample2.csv" file:

    import pandas as pd
import mplfinance as mpf
import matplotlib.dates as mdates

trading_pair= "DOTBUSD"

df = pd.read_csv('sample2.csv', index_col = 'Start Date', parse_dates = True)

# Plotting
# Create my own `marketcolors` style:
mc = mpf.make_marketcolors(up = '#2fc71e', down = '#ed2f1a', inherit = True)
# Create my own `MatPlotFinance` style:
s  = mpf.make_mpf_style(base_mpl_style = ['bmh', 'dark_background'], marketcolors = mc, y_on_right = True)    

# Plot it
trading_plot, axlist = mpf.plot(df,
                    figratio = (10, 6),
                    type = "candle",
                    style = s,
                    tight_layout = True,
                    datetime_format = '%H:%M',
                    ylabel = "Precio ($)",
                    returnfig = True,
                    # Activamos nontrading = True
                    show_nontrading = True
                    )
# Add Title
symbol = trading_pair.replace("BUSD","") + "/" + "BUSD"
axlist[0].set_title(f"{symbol} - 3m", fontsize = 25, style = 'italic', fontfamily = 'fantasy')

# Elejimos los minutos que serán los "ticks" principales
major_ticks = [45,51,0,9,15,21,30,39]

# Configuramos los "tics" principales
axlist[0].xaxis.set_major_locator(mdates.MinuteLocator(byminute=major_ticks))

# Configuramos los "ticks" secundarios cada 3 minutos
axlist[0].xaxis.set_minor_locator(mdates.MinuteLocator(range(0,60,3)))

# Configuramos la cuadrícula
axlist[0].grid(visible = True, axis = 'y', which='both')
axlist[0].grid(visible = True,  which='minor')

    This returns the following graph:

    Graph

    • 1