How to display single points with line trace?

đź“Š Plotly Python
1

Hi Plotly community,

As this is my first message I’d like to say big thanks to Plotly developers for the good work :slight_smile:

I have a question: I use a line trace (see minimal example). At the middle, there is a single point which is not displayed.

My goal is to display lines without markers, and single points with markers.

If I enable `“mode”: “lines+markers”, there is a marker for every point. I would like to render markers only for single points not connected to any line, because with a very large number of points, the chart is too heavy visually if every point has a marker.

Is there any option allowing me to achieve what I want? Or should I use two traces, one with lines and the other with markers (for single points), and pre-process my data first? (something like this example)

2

Does the data belong to the same graph? Or is it one graph with both continuous lines and points? Or is it more like the line is a fit of a Scatter plot? I’m a little confused. Because in that case I would use separate traces. Can you give an image of what it should look like?

3

Thanks @klnrdknt

Data is one time series in my case. Could be like:

[
  ["2000", 10],
  ["2001", 11],
  ["2002", null],
  ["2003", 15],
  ["2004", null],
]

To be more concrete here is an image of the expected result:

image
image1562Ă—764 17.8 KB

The user should feed like it’s a single time series.

I would like:

  • uninterrupted successions of points to be rendered as lines without markers (like the 2 first points of the list)
  • single points (like ["2003", 15]) to be rendered as markers

Maybe I’m missing some trivial point. I’m open to other approaches.

4

For the record, I answer myself because I found a solution.

{
    type: "scattergl",
    x: x,
    y: y,
    mode: "lines+markers",
    marker: {
        size: 3,
    },
}

In my first post I eliminated the "mode": "lines+markers" solution because markers were too big, but I discovered that marker size can be modified.

Note: I used “scattergl” for performance reasons because x and y are large.

Now I can see markers and lines:

image