Statistical Trend Lines

Use this when you want to add a best-fit trend line alongside your raw data.

Regression — DF_REGRESSION

Fit a linear or curved trend line to your data and add the predicted values as a new column.

Use this when: you want to visualise the underlying trend in noisy data, or project where a value is heading.

Available types: linear (straight trend line), polynomial (curved trend line).

Before

Code
[
  { "day": 1, "kwh": 12 },
  { "day": 2, "kwh": 14 },
  { "day": 3, "kwh": 11 },
  { "day": 4, "kwh": 16 },
  { "day": 5, "kwh": 15 }
]

After — adding a linear trend column

Code
[
  { "day": 1, "kwh": 12, "predicted_kwh": 11.6 },
  { "day": 2, "kwh": 14, "predicted_kwh": 12.7 },
  { "day": 3, "kwh": 11, "predicted_kwh": 13.8 },
  { "day": 4, "kwh": 16, "predicted_kwh": 14.9 },
  { "day": 5, "kwh": 15, "predicted_kwh": 16.0 }
]

Code
{
  "type": "DF_REGRESSION",
  "config": {
    "xField": "day",
    "yField": "kwh",
    "type": "linear",
    "outputField": "predicted_kwh"
  }
}

For a curved trend, use "type": "polynomial" and optionally set "degree" (default is 2).