[Bug] Using dataset with random numbers is mush slower

[zkycaesar]

Version

5.5.1

Link to Minimal Reproduction

https://echarts.apache.org/examples/zh/editor.html?code=MYewdgzgLgBADlA-mArgWxgXhgRgAwEEDcAUADYCmsAHgGoCGZKFEWMYFA7jAJJhQBmAEwBBAE5j6ATwAUCZOgCUpSrCkMmLNh24AxMiHpQAbABZxk2fNRplJAPT2YqmOsbMIQ7Vxj7Dg0QlpOSQbOwAzEDEYGRcASzY8IhgEgB54UPRkuIBqHMUYAG8SGFKYOncWAG04gF02ONIymEdXDQ8a-uwAWSMACwA6STAAExA0GTtmt00ITrZeqEGIOLAZBKd8AjsAXxKykhAEOPA2YtLqEWo4iAAuIv3SqCk4CnuAcgA3SveAGn2dv9SlIrjd7ucys9Xh9vpp3gCgTARkZ6BAqODHmU4iMPlBJMAKABaHB_TGlEZxNAUSAnSC3Kpk5pFdj0Kkfah_GBQt4wd6rKDvQEwRnNQostm8qSc7kfcIGIyCxm1RHNCAgFBiAn3KoVWa_NqVCC1BH7NFiOIsDFM8U8954-gE4mk60y3lkVYUZ1MihoOB9VFgh7W5oUiD0ABGlBxXLEzBVZUBjNaYd97rAAHMPmQoFBw17mmmKABlZ6UcEwTjYpb3HAwRPWiB9ECcItSNDhkBke7hRho-PkynUlbgO4M4NlMVgVm2jn6118_iC_Uiic2j5SucvW1y_yK63KxnU0AjHkQ8fvKXr_PB94c9nw631plmi0QAAy0nVUAAQlIPmJmy9PYdiIIA

Steps to Reproduce

I find that if I use random numbers as dataset, like Math.random(), the line chart rendering will be very slow. But using dataset with some pattern, like Math.sin(i), it will be much faster.

Please have a look at the demo: demo. Uncomment line 7, and comment line 8.

Am I doing something wrong or this is a performance issue?

Current Behavior

Using random numbers as dataset will be much slower.

Expected Behavior

The performance should be the same.

Environment

- OS:
- Browser:
- Framework:

Any additional comments?

No response

[zkycaesar]

如果使用随机噪声作为数据，在4k显示器下，即使只有3000个点，axisPointer也会变得很卡。请看demo

这个性能感觉不太能接受。想问下是否有什么选项可以优化这个性能？

[zkycaesar]

把线宽调低，可以缓解这一性能问题，先这样吧。大概是像素点越多渲染约慢。
实测下来，uplot和lightningchart在折线图的渲染性能这方面要比echarts强不少，如果确实有性能需求，可以考虑这俩。