Adding heat element to a XY (scatter graph)

[RT705]

Hi

I'm trying (without success) to find a way in Excel to enhance the basic scattergraph below. There's a very high density of points in the bottom left hand corner which means they are multiply overlapping and don't give a good visualization of the weighting of this part of the data.

Is it possible to show a heat type map with this - i.e. the areas with most overlapping come up with say red, then yellow for medium density of scatterpoints and then green or blue for the least dense?

Thank you for any help with this

Rob

 

[vletm]

RT705
Many thing are possible with a sample file and with clear visual image of Your needs.

 

[NARAYANK991]

Hi ,

Beyond just putting the above points in different segments , how will just adding colours such as red , orange , yellow ,... add value or explanation to the above graph ?

What is it that you want the above chart to convey ?

Narayan

 

[p45cal]

If I could jump in here NARAYANK991, in the bottom left of the chart, where a blue dot/blob is visible, it shows the same whether there are 10 blue blobs beneath it or not, so I can see the value in changing the colour/intensity of colour to indicate that there are several of them in one place, or at least close to each other. That said, there are other ways to show clustering (I just haven't thought hard enough yet how to do it!).
A quick solution might be to make the points semi-transparent, that way points on on top of each other or overlapping will show an increase in colour saturation.
RT705, if you were to provide a workbook with just the values and the chart we could have a play with it (saves us the bother of recreating your chart) to try and find ways to convey the information.

 

[NARAYANK991]

If I could jump in here NARAYANK991, in the bottom left of the chart, where a blue dot/blob is visible, it shows the same whether there are 10 blue blobs beneath it or not, so I can see the value in changing the colour/intensity of colour to indicate that there are several of them in one place, or at least close to each other.

Hi ,

Agreed , but only the OP can explain what he wants to convey.

Many points may have the same (x,y) value ; also many points may have (x,y) values closely clustered together ; are these two situations to be treated differently or in the same manner ?

Unless it is known what the X-axis and Y-axis values represent , it is difficult to say what the chart is supposed to convey.

Narayan

 

[p45cal]

After 2 weeks it begins to look like @RT705 has lost interest…