Real-time charting resource list

[goodboy]

Well this is a super old issue that i'm now updating :joy:

piker is being built to eventually support multiple "front ends" for charting - that's a much longer term and obviously ambitious task but it's worth stating up front so that new users or contribs realize that we aren't locked into any particular UI framework or methodology per say.

We've obviously longed pyqtgraph atm, so i don't think i need to write up too on that front :joy:

---

General web searched research about what the peeps think / say..

• reddit discussion on plotting quote streams
• a link to a talk (from an old pal see comments) https://speakerdeck.com/jakevdp/pythons-visualization-landscape-pycon-2017
• py plotting resource list: https://github.com/batiron/datavizstash/blob/37aa90a511e4df67f3ae65e3490581642a3261b6/datavizstash.md#python-tools

---

Actual frameworks worth keeping an eye on:

• the vispy realtime signals example (which I can't get running) as well as the notebook tut
• https://github.com/blacksmith-capital/likhtar
• http://rrag.github.io/react-stockcharts/

immediate mode guis with already built-in charting frameworks:

• https://github.com/hoffstadt/DearPyGui
• https://github.com/emilk/egui (from rust land :sunglasses:):
  • plotting apis: https://docs.rs/egui/latest/egui/?search=plot

[goodboy]

@dkeitel made a good point to checkout the visualizations ecosystem talk before diving too deep in any one lib.

[goodboy]

PyQTGraph one of the collaborator projects behind vispy is worth checking for sure.

[goodboy]

An example of real time chart updates using bokeh and iex.

[goodboy]

Another list of python plotting resources.

[goodboy]

After getting frustrated with the new version of webull which now is only supported on chrome I started digging into QT's built in candlestick charting so see how quickly I could get something going.

Here's a small list of resources I've found in this little endeavour.

Candlestick charting:

• the Qt Charts overview
• the candlestick charting tutorial
• the candlestick series QML type
• stack overflow answer using PyQt5 (I ended up porting this to PySide2 - I'll push that in a branch)
• chart zooming using a QRubberBand
• PySide2 candlestick api
• a financial charts demo using Qcustomplot
• SO answer on plotting candlesticks from a pandas.DataFrame

Theme-ing

• dark style sheet
• Qt5 dark palette gitst 1, gist 2 (second is a link from first)

Miscellaneous Qt stuff

• Qt for Python tutorials
• SO answer on killing PyQt on ctr-c
• a vispy issue describing how to best integrate their canvas into an existing Qt application (this is likely where we want to go down the road to get nice GPU fuelled performance though would require writing our own OpenGL candlestick geometries)
• blog post on PyQt5 vs. PySide2
• pyqtgraph github repo overview of wrappers for Qt
• quamash issue on how to get PySide2 working

[goodboy]

Started poking around in pyqtgraph and found some already working examples:

• the custom graphics example renders a candlestick chart
• this dude's question on the google group
• this other dude's example of real-time charting using PyQt5 on groups.io where the author of ib_insync pipes in
• another thread from the ib_insync group discussing Qt5 integration

[goodboy]

Here's some more OpenGL related projects:

• ponzi2
• pyGLFW

[goodboy]

On the note of using QCustomPlot (a C++ lib) there is support for plotting bar style OHLC points using the QCPFinancial plottable object. It might be worth digging into the C++ lib to see if we can extract that bit for use.

[goodboy]

Cool web-based (js) libs list:

• likhtar
• react stock charts

[goodboy]

DearPyGui is another C++ backed lib that runs on OpenGL3 for Linux. Seems to have support for real-time highly interactive plotting.