问题
I have a use case where I pull and plot Forex data in the form of ask and bid on the graph and this is based on minute, hour or day candlesticks and I am only plotting the closing value for bid and ask as a spline.
Issue
The endpoints that I am using their Forex data feed is not available from Friday 8pm to Sunday 9pm. Let's say I have the following use case
- Plot a 200 data points (i.e. hourly candle) on the graph at a time, on new candle phase out the oldest and plot the newest
- The X-Axis is the data point's
datetimeinUTC ISO Format - The Y-Axis is the value of ask and bid
- At a time the graph shows
4.1 days worth of hourly candle
The graph is pretty and smooth when the datetime on x-axis do not have any gaps. The candle timestamps are returned in the following form
2018-06-29T16:00:00.000000000Z
2018-06-29T17:00:00.000000000Z
2018-06-29T19:00:00.000000000Z
2018-06-29T20:00:00.000000000Z (Friday last candle 8pm)
2018-07-01T21:00:00.000000000Z (Sunday first candle 9pm)
2018-07-01T22:00:00.000000000Z
2018-07-01T23:00:00.000000000Z
2018-07-02T00:00:00.000000000Z
As soon as there is a gap (around 47-49 hours) everything gets compressed to accommodate for the stretched part of the graph as shown below (Using plotly.js)
Mostly, financial graphs do not show the gap. For example, looking at the TradingView interactive chart you will see the following for weekend
If you know how to tackle this issue with open source charting libraries (excluding TradingView) then please share your solution.
回答1:
TL;DR
Based on my findings, you can exclude the gaps using HighStock of Highcharts library. Highcharts is open source and only free to use for Non-commercial use, for more details see highcharts licensing
plotly.js
I initially started off with plotly.js because it really is a very good charting library to be honest. After I came across the datetime gap issue (see screenshot in question) on x-axis I researched for solution and came across the following issues
https://community.plot.ly/t/tickformat-without-weekends/1286/5 (forum)
https://github.com/plotly/plotly.js/issues/1382 (locked github thread)
https://community.plot.ly/t/x-axis-without-weekends/1091/2
As per the Github issue link above, this is not achievable at the moment for how the datetime is handled on x-axis. There are workaround (like 3rd link above) but, I tried some and it either didn't work or I lost some piece of information.
I came to the conclusion that even though plotly is a great library, it simply cannot help me in this case. If you think it does, please share your solution :)
C3.js
After some more research of charting libraries, I give C3.js a try and after editing their live demo example with datetime for x-axis I observed the same weekend gap issue (see below)
I did not bother researching why and that's probably because of being based on D3.js just like plotly.js. If you think the datetime gap can be handled with C3.js then please feel free to share your solution.
Highcharts
I finally came across the family of Highxxxxx charting library and at first look it looked so pretty and offered all the functionalities that I wanted. I tried Highchart and I got the same weekend gap issue
Well, I started to think that the only solution is to NOT TO PLOT the datetime in x-axis at all and somehow show the datetime in tooltip and that seem to do the job as shown below
But, being too stubborn to give up I continued my research. I really wanted to get the charting done like TradingView and other stock visualization.
I decided to do some more research and little did I knew that I was so so close to a solution by the very same charting library Highcharts and it is called HighStock
Highstock of Highcharts
HighStock by default takes care of the gaps as shown below
I used the following piece of Javascript
Highcharts.setOptions({
global: {
useUTC: false
}
});
Highcharts.StockChart('chart', {
chart: {
type: 'spline',
events: {
load: function () {
console.log('Highcharts is Loaded! Creating a reference to Chart dom.')
//store a reference to the chart to update the data series
var index=$("#chart").data('highchartsChart');
highchartRef = Highcharts.charts[index]; //global var
//Initiating the data pull here
}
}
},
title: {
text: 'Chart title'
},
series: [
{
name: 'Ask',
data: [],
tooltip: {
valueDecimals: 7
},
marker: {
enabled: true
}
},
{
name: 'Bid',
data: [],
tooltip: {
valueDecimals: 7
},
marker: {
enabled: true
}
}
]
});
There you have it, a gapless (not sure if it is a real word) financial chart with datetime x-axis for financial data.
Libraries I did not try
I have researched but, did not try the following charting libraries
- Google charts (didn't feel right for unknown reasons)
- Chartist (didn't see much of interactivity and financial related chart types and examples
- Chartjs (same as Chartist)
- TradingView (I give up after reading their terms of use - weird clause, my point of view)
- A majority of others that lacked the level of documentation, interactivity, finance-related charts like
PlotlyandHighcharts.
Final thought
I found Highcharts to be
- Well documented
- A ton of jsfiddles examples around
- Lots of useful SO answers
- Pretty and very functional
- Many options
- Easy to use
Similar SO questions relating to gaps
How can I hide data gaps in plotly?
HighCharts datetime xAxis without missing values (weekends) (Highcharts)
Remove weekend gaps in gnuplot for candlestick chart (candlestick-chart, gunplot)
How to remove gaps in C3 timeseries chart? (c3/d3 - fiddles are not working so its hard to see the solution)
How to remove weekends dates from x axis in dc js chart (c3.js - the gap is some gap)
回答2:
Using category as the xaxis type worked for me quite nicely (notice the go.Layout object):
import plotly.graph_objects as go
layout = go.Layout(
xaxis=dict(
type='category',
)
)
fig = go.Figure(data=[go.Candlestick(x=self.price_df.index.to_series(),
open=self.price_df[e.OHLCV.OPEN.value],
high=self.price_df[e.OHLCV.HIGH.value],
low=self.price_df[e.OHLCV.LOW.value],
close=self.price_df[e.OHLCV.CLOSE.value])],
layout=layout)
fig.show()
回答3:
According to documentation
try to add this before fig.show()
fig.update_xaxes(
rangebreaks=[
dict(bounds=["sat", "sun"]), #hide weekends
]
)
or
fig.layout.xaxis.type = 'category'
回答4:
Plotly is now supporting to hide weekend and holidays in chart with help of rangebreaks. check this link:https://plotly.com/python/time-series/ go to the section:"Hiding Weekends and Holidays"
Below is section copied from this link, example is for plotly.express but it also works for plotly.graph_objects
Hiding Weekends and Holidays The rangebreaks attribute available on x- and y-axes of type date can be used to hide certain time-periods. In the example below, we show two plots: one in default mode to show gaps in the data, and one where we hide weekends and holidays to show an uninterrupted trading history. Note the smaller gaps between the grid lines for December 21 and January 4, where holidays were removed. Check out the reference for more options: https://plotly.com/python/reference/#layout-xaxis-rangebreaks
import plotly.express as px
import pandas as pd
df = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/master/finance-charts-apple.csv')
fig = px.scatter(df, x='Date', y='AAPL.High', range_x=['2015-12-01', '2016-01-15'],
title="Default Display with Gaps")
fig.show()
Default Display with Gaps
import plotly.express as px
import pandas as pd
df = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/master/finance-charts-apple.csv')
fig = px.scatter(df, x='Date', y='AAPL.High', range_x=['2015-12-01', '2016-01-15'],
title="Hide Gaps with rangebreaks")
fig.update_xaxes(
rangebreaks=[
dict(bounds=["sat", "mon"]), #hide weekends
dict(values=["2015-12-25", "2016-01-01"]) # hide Christmas and New Year's
]
)
fig.show()
来源：https://stackoverflow.com/questions/51180269/how-to-remove-weekend-datetime-gaps-from-x-axis-of-a-financial-chart