Description:
I need a tool that allows me to import candlestick time series data, display it and allow me to create labels for it. The candlestick data can fit in ram or be read form file
Similar Tools:
https://labelstud.io/playground/ allows you to create labels for time series data but it's not candlestick and it takes the X/time axis as the label instead of arbitrary y axis values.
https://github.com/tradingview/lightweight-charts already has all the candle drawing and labeling functionality but isn't set up to load random candles and save the labels.
Inputs:
candlestick data from file on disk (can be user input for path), "Number of Candles" user input, "Next" button
Dataset is a table and can be reformatted to fit ease-of-development but has this basic shape:
[
{
"timestamp": datetime
"open": float
"high": float
"low": float
"close": float
},
...
]
Workflow:
Display a random sequence of "Number of Candles" length worth of ohlc candles from dataset (using preferably tradingview's charting library) . Allow the drawing of any number of horizontal lines and save their Y axis value as output. When hitting Next, we save the dataset or append to it in a file as described in "outputs" section and redraw the chart as another random sequence of candles.
A very poor quality mockup is attached.
Outputs:
{
"inputs": {
[
{ # Input can be reformatted to fit data ingestion needs, but it is basically a table
"timestamp": datetime
"open": float
"high": float
"low": float
"close": float
},
...
],
"labels": [
<float>,
....
]
}
Other Considerations:
The project may be open sourced on your personal github repo if desired.
