Line chart
Dashboard component reference for the Line chart, its series, axes and stacked plots, window and retention, drawing fidelity and the toolbox.
The Line chart follows points over time: each series reads one Logic address, sampled on the chart's own cadence into the station's live-series collector, so the chart opens onto a filled window instead of an empty axis. The same drawing serves a run's recorded chart in Process and Histories; what a dashboard adds is where the configuration comes from — this component's own properties.
Content
| Property | What it does | Values | Default |
|---|---|---|---|
| Chart title | Optional heading drawn inside the plot. This is the chart's own heading, not the component's corner title (that one lives in the shared Appearance section). | text | empty |
Series
One card per series, in the order declared. Add series appends one; each card carries:
| Property | What it does | Values | Default |
|---|---|---|---|
| Address | Rooted Logic address the series reads. The field suggests from the live point list, filtered to numeric points, and shows the point's current reading once named. | address | empty |
| Name | What the legend and the tooltip call the series. | text | blank falls back to the address |
| Color | The line's ink as #RRGGBB. |
hex color | blank takes the palette's turn |
| Decimals | How many places this series' readings are written with under the cursor. Per series, because the channels of one chart are not read to the same precision — a pressure to two places beside a count that has none. | 0 to 15 | blank uses the station's preference |
| Join the line across gaps | What the line does where this series was not read. Off, the line breaks at the hole — the honest picture, because the station has nothing for that stretch. On, it is carried straight across, which is how a setpoint that simply held between writes is read. | on / off | off |
| Plot | Which of the chart's stacked plots the series is drawn in. Only asked while the chart has plots under its own. | the chart's own · a stacked plot | the chart's own |
| Axis | Which value axis the series scales against — first or second. Only asked inside the chart's own plot while a second axis exists; a series in a stacked plot has that plot's single axis. | first · second | first |
| Draw this series | Whether the series reaches the canvas at all. Persisted, so a channel left out of the picture stays out the next time the screen opens — the toolbox's show/hide gesture is for looking, this is for authoring. | on / off | on |
Chart
| Property | What it does | Values | Default |
|---|---|---|---|
| Shape | The shape every series takes over the axis: a continuous measurement is a line, a total under it is an area, a value that holds until written again is a step, a per-interval quantity is bars over time. | Line · Area · Step · Bars over time | Line |
| Legend | Where the legend sits. | over the plot · under the plot · none | over the plot |
| Navigation | How far the operator may walk the time axis: a fixed frame that nothing moves, wheel and drag over the plot, or wheel and drag plus a navigation bar under the plot showing where the frame sits. | fixed frame · wheel and drag · wheel and a navigation bar | wheel and drag |
Axes
Every axis is a card: a name, a unit (appended to the name in brackets), a division count and, for value axes, held ends and an edge.
Time axis
| Property | What it does | Values | Default |
|---|---|---|---|
| Counts in | How the axis presents sample positions: elapsed time relative to the window's edge, wall-clock timestamps (formatted per the station's localization), or the plain sample number. | time since the window opened · the clock · the sample number | time since the window opened |
| Name / Unit | The axis' name and unit. | text | empty |
| Ticks | How many divisions the axis is asked for. A stated count is a hint the engine honours as closely as round values allow, not a promise of that many ticks. | 2 to 20 | blank (auto) |
Value axis (and the optional Second value axis, for series whose scale has nothing to do with the first — without one, every series shares the only scale there is)
| Property | What it does | Values | Default |
|---|---|---|---|
| Name / Unit / Ticks | As on the time axis. | ||
| Min / Max | The ends the scale is held at. Left to itself a value axis re-fits at every sample, so the tallest line is always at the top of the frame and nothing can be compared with a minute ago; holding an end stops it. Each end is held on its own — a floor alone keeps the zero line while the ceiling breathes. | numbers | blank (fit) |
| Drawn on the | Which edge of the plot the ruler stands on. | left · right | left |
Stacked plots
Two quantities read at the same instant do not have to share one ruler: a plot stacked under the chart's own gets a scale of its own and the same time axis, so temperature and flow are read one under the other instead of squeezed onto each other's ruler. A chart is at most four plots tall — its own plus up to three stacked under it; each stacked plot's card carries its axis (name, unit, ticks, held ends, edge) and:
| Property | What it does | Values | Default |
|---|---|---|---|
| Height | How tall the plot is drawn against the chart's own, which is 1. | 0.25 to 4 | 1 |
A series states which plot it lands on in its own card.
Evaluation
| Property | What it does | Values | Default |
|---|---|---|---|
| Draw the approved range of the run under way | Draws, behind the series, the approved band of the run the unit is executing right now. The limits come from the evaluation of whichever procedure the unit is running, and a limit that follows a reference channel moves with it. It names no unit of its own: the unit is the one the series already read — on a model's screen it arrives when a unit opens the screen, and a screen reading no unit has no run to follow. | on / off | off |
This is how a screen chart follows a running procedure; a chart of a recorded run in Histories already carries its own run's bands.
History
Two different knobs: the window is what the chart frames, the retention is what the station keeps behind it — a chart may keep ten minutes and frame the last one, so navigating back walks what was kept instead of meeting an empty axis. Samples live in memory; restarting the application starts the collection over.
| Property | What it does | Values | Default |
|---|---|---|---|
| Show | What the frame opens on: a rolling window that follows the newest sample, or everything the station kept for the chart, however far back that goes. | a rolling window · everything kept | a rolling window |
| Window (s) | The stretch a rolling frame shows. Only asked while the frame rolls. | 5 to 86,400 (1 day) | 60 |
| Sample every (ms) | The chart's own sampling cadence. Changing it starts a new buffer from now, because a cadence is part of a collected series' identity. | 100 (the sweep) to 3,600,000 (1 h) | 500 |
| Keep | What the retention numbers mean. | just the window · a stretch of time · a number of samples | just the window |
| Keep for (s) | The stretch kept per series while retention is a stretch of time. | 5 to 604,800 (7 days) | 600 |
| Keep (samples) | The count kept per series while retention is a number of samples. | 2 to 20,000 | 2,000 |
The station grants retention inside three ceilings: 20,000 samples per point, 60,000 per component and 200,000 for the station as a whole. A retention a ceiling could not grant whole is said under the chart itself, not only in the editor. When the station's own ceiling is already spent, a new series collects nothing at all and says so: "The station is already keeping all the history it allows, so this series collects nothing. Shorten the retention of another chart to make room.".
Fidelity and performance
How much of the data reaches the canvas — the chart's most consequential setting. Drawing every sample stalls as the channels and the window grow; reducing without saying so is worse. The reduction happens at draw time and never in what is stored, and it is adaptive: the chart counts the points inside the visible range, and below the threshold the reduction disappears — zooming in always reveals the real data.
A preset fills every field at once; touching any single field makes the set Custom by itself, because nothing is stored beside the values.
| Preset | What it writes |
|---|---|
| Maximum fidelity | Threshold 100,000 (past what any canvas resolves — never reduce), keep the peaks, symbols up to 1,000 points, draw outside the frame, no animation, mode shown in the corner. |
| Balanced | The values a chart is born with (below). |
| Maximum performance | Threshold 400, average each interval, no symbols, nothing drawn outside the frame, no animation, no mode mark. |
| Custom | Not a choice: the reading shown when the fields match none of the named sets. |
| Property | What it does | Values | Default |
|---|---|---|---|
| Draw every sample up to | How many points inside the visible range the reduction switches off below. It counts the whole picture — every drawn series together. | 100 to 100,000 | 2,000 |
| Above it, keep | Which reduction runs above the threshold, named by what it preserves: the minimum and maximum of every interval (a peak that lasted one sample cannot be dropped), the point that keeps the curve's silhouette, or the mean of each interval — the cheapest, and the one that flattens a peak. | Preserve peaks · Preserve shape · Average | Preserve peaks |
| One symbol per sample up to | How many visible points may still each carry a symbol; zero never draws them. A symbol only means anything while the reduction is off. | 0 to 5,000 | 200 |
| Draw what lies outside the frame | Whether what is outside the visible range is drawn at all (coarsely — it becomes the real data again the moment it is framed). | on / off | on |
| Animate the drawing | Whether the drawing animates. Off, because live data that slides is harder to read. | on / off | off |
| Say in the corner which mode the chart is in | A discreet corner mark saying whether the screen shows the measurement or a reduction of it. Clutter on an operating screen, essential on an analysis screen. | on / off | off |
An estimate under the fields says what the configuration costs, live as it is typed.
Toolbox
The Toolbox section picks which buttons the operator gets on this chart, so an operating screen where the operator must not move the scale stops carrying an analysis screen's bar. With none ticked, no bar is drawn at all.
| Tool | What it does | In the default set |
|---|---|---|
| Zoom into an area | Arms box-select zoom over the plot. | yes |
| Reset the zoom | Returns the frame to the whole of what is plotted. | yes |
| Set the scale by hand | Pins the time period and the value range by hand, or hands either back to the fit. The button stays lit while anything is pinned. A pinned frame holding no samples says so under the drawing instead of reading as a chart that stopped. | no |
| Read the window's minimum, average and maximum | Opens a panel with each series' lowest, mean and highest value inside the window shown. Measured only while the panel is open. | no |
| Draw reference lines | Draws a line at a value or at an instant. The lines live only as long as the screen stays open — nothing is persisted — and the button stays lit while any line is drawn. | no |
| Save as PNG | Downloads the drawing as a PNG. | yes |
| Copy as PNG | Puts the drawing on the clipboard as a PNG. | yes |
| Show or hide series | Shows or hides one series without editing the screen. | yes |
| Turn the tooltip on and off | Toggles the hover tooltip. | yes |
| Full screen | Fills the window with this chart. | yes |
On a model's screen
On a screen defined on a Process model, the series address fields lead with the model's abstract channels: picking one binds the channel and leaves the unit to be answered when a unit opens the screen. The station's own points keep their place behind them — a model's screen may legitimately read one.