STEP

step line (values hold between changes). Needs: ordered periods + series whose value HOLDS until it changes (rates, stock, policy).

change-over-timetrend

DL-01 AGRAW

Use when

levels that hold between discrete changes: stock, rates, headcount, policies
the timing of the change is the message

Do not use when

continuously varying measures (use CH-TIM-01)
cumulative progress (CH-TIM-02)

See instead

values change continuously: CH-TIM-01

Alternatives

CH-TIM-01 (continuous change)

Rules

horizontal-then-vertical interpolation; the renderer draws the hold explicitly
y starts at zero
series labeled at line ends, never a legend
annotation marks the change that matters

Never do this

smooth or diagonal interpolation of held values (invents data)
markers on every step
legends

QA before delivering

step interpolation verified
y at zero
no legend
title states the level change with a number
source present

Data shape

title	action title with a number
eyebrow	caps, DL chrome only
source	bottom-left always
unit	on top tick
x	ordered periods
series	one key + optional context, values hold between points
annotation	optional, anchored at a data point

Sample data (golden)

{
  "title": "Stock held at 41 units for three weeks after the June draw-down",
  "eyebrow": "WAREHOUSE STOCK . W20-W28",
  "source": "Source: sample dataset v1, 2026-07-12",
  "unit": " units",
  "x": [
    "W20",
    "W21",
    "W22",
    "W23",
    "W24",
    "W25",
    "W26",
    "W27",
    "W28"
  ],
  "series": [
    {
      "name": "Stock",
      "values": [
        64,
        64,
        64,
        58,
        58,
        49,
        41,
        41,
        41
      ],
      "role": "key"
    }
  ],
  "annotation": {
    "text": "June draw-down: -23 in three weeks",
    "series": "Stock",
    "index": 5,
    "tx": 470,
    "ty": 200,
    "anchor": "start"
  }
}