18 replies to this topic

[Hans van der Maarel]

Inspired by the discussion on new flow maps I decided to try something similar to what Dennis had done, albeit in a more flowing shape. Also used some real data, courtesy of the CBS (Central Bureau of Statistics), just for the sake of it.

So this is what I came up with:
 flow_patterns_nl.png   30.75K   165 downloads

Please let me know what you all think of this. The graph axes show the total amount of incoming/outgoing migration, circle sizes represent the migration within each region. The width of the flow lines indicates how many people migrated from A to B (line color = region they moved out of).

[rudy]

Inspired by the discussion on new flow maps I decided to try something similar to what Dennis had done, albeit in a more flowing shape. Also used some real data, courtesy of the CBS (Central Bureau of Statistics), just for the sake of it.

So this is what I came up with:
 flow_patterns_nl.png   30.75K   165 downloads

Please let me know what you all think of this. The graph axes show the total amount of incoming/outgoing migration, circle sizes represent the migration within each region. The width of the flow lines indicates how many people migrated from A to B (line color = region they moved out of).

Looks good but I had to read your posting to figure out what the width of the flow lines represented (i.e. in or out migration). Perhaps you could add a note or legend item to that effect so it is immediately clear.

How would it looks a geographic representation?

[BioGeoMan]

I think it makes perfect sense and a great way to show three different data sets in one graphic. Looks like those northerners don't migrate much...interregionally or intraregionally.

One note: you may want to define your units, even though it is assumed to be humans...it could be dairy cows!!

Thanks Hans

[Hans van der Maarel]

Thanks for your comments. I kinda figured the line color would be obvious enough, but I've added the explanantion. Also added a little outline to overlapping lines to separate them a bit more.

 flow_patterns_nl2.png   34.03K   81 downloads

The north is pretty sparsely populated (well, for Dutch standards anyway) so the absolute numbers are a lot less. To put it in perspective, there's about as many people living in region North as there are in Rotterdam and its aglommeration. On top of that, the 4 biggest cities in the country are all in region West.

Rudy,

A geographic version would be something like my option 2 in the original discussion that inspired all this. However, I used random numbers for that one and as it turned out my regional division was a bit off as well.

[patdunlavey]

I was initially confused by the graph, not noticing that the x/y positions of the circles were determined by their inflow/outflow values. I expected to have the positions correspond to something approximating a centroid of the geographic region in question. The thing is, positioning the circles based on their inflow/outflow totals is redundant, since the flow lines communicate these inflow/outflow values exceptionally well. It's intuitively clear which flow lines are inflow and which are outflow by their color, and likewise their width gives a highly intuitive sense of their magnitude. I think I would position the circles as geographic centroids, possibly superimposed on the map of the regions itself, and connect the flow lines as you have done.

-Pat

[Hans van der Maarel]

Pat,

Yes, I was thinking the same thing re. the axes. Also, since the center of the circle corresponds with the position on the graph, it's a bit hard to read.

Will try a geographic version next. It's a fun thing to experiment with anyway.

[Rick Dey]

Hans,

Maybe I'm misunderstanding it, but shouldn't Intraregional be the movement within the regions and Interregional be between?

[DaveB]

As it stands it's a graph, not a map (not that there's anything wrong with that!). The idea of placing the circles on a map sounds good to me.

One thing that I keep thinking when I look at the graph is that a lot of people are leaving the east to go to the north, but that may just be due to the prominence (width and length) of the green flow line. Maybe if you moved the thicker lines to the back? Arranging the circles geographically might also help with this and with determining geographical patterns more readily.

[Hans van der Maarel]

Hans,

Maybe I'm misunderstanding it, but shouldn't Intraregional be the movement within the regions and Interregional be between?

I'll have to look that up...

[MapMedia]

Nice job Hans! I like the representation IMHO. I was thinking it wold be useful to replace interregional (circles) with immigration to the Netherlands, and the lines as interregional (between regions). As for intraregional, that would be within each region, which is not shown. That would show people coming into the Netherlands and how people move within the country within 1 year.
Would be a great submission to a magazine or newspaper.

-Chris

[Hans van der Maarel]

Nice job Hans! I like the representation IMHO. I was thinking it wold be useful to replace interregional (circles) with immigration to the Netherlands, and the lines as interregional (between regions). As for intraregional, that would be within each region, which is not shown. That would show people coming into the Netherlands and how people move within the country within 1 year.
Would be a great submission to a magazine or newspaper.

Actually, the sizes of the circles represent how many people have moved within that region.

One of the reasons why I plotted them like this (with the 2 axes) was so that it would be obvious which regions have a discrepancy in immi-/emigration, but it turns out they're all pretty much balanced in that respect.

I was also amazed at the fairly low numbers for migration between regions (you all notice I'm staying away from the long, difficult words now). South to North for example is merely 1800-something (makes sense, it's much cooler down here...)

[CHART]

The added explanation on your second map did the trick for me. I understood the graph right away.

Maybe a white outline on the circle would look better. I would enlarge the map a little and add a few major cities. This could help explain some of the migration patterns... e.g. maybe the presence of large cities = more intraregional migration?

Anyway nice geo-graphic representation.

What is the destination media?

I think when posting in the map gallery one should always note the destination media, the size, the audience, time frame to do the work. In sum a list that can help others comment properly your work.

cheers.

[MapMedia]

Great. One last idea, is to combine the legend that you have on the right, into the graphic itself: put labels on the flow lines and in the middle of the circles. Then the graphic would be cleaner and more direct.
The colors are very good too. I really enjoy these sort of pseudo map graphics too.

Bubble maps are so common now in the media - SEE NYT OLYMPIC MAP - goes to show ppl have no problem interpreting this sort of representation, given the assumption the readership has a basic grasp of where these counties should be.

Attached Files

•  Olympic_map.png   76.97K   49 downloads

[Hans van der Maarel]

And here's a geographic version.

 flow_patterns_nl_geo.jpg   105.63K   84 downloads

@Chart
Since this is a 'personal hobby' project, there isn't really a target audience or media, other than Cartotalk. I intended for this project to be a good topic of discussion here, as well as a learning-moment for myself.

[Green Palolo]

And here's a geographic version.

 flow_patterns_nl_geo.jpg   105.63K   84 downloads

@Chart
Since this is a 'personal hobby' project, there isn't really a target audience or media, other than Cartotalk. I intended for this project to be a good topic of discussion here, as well as a learning-moment for myself.

I would vote for the geographic version, for me it communicates the data in a much better fashion.