GMTdataset to/from DataFrame in Julia

Q&A GMT.jl Q&A
1

I have loaded a shapefile as so:

bldg = gmtread(joinpath(directory, "bldgs_preprocs_E4.shp"))

Vector{GMTdataset} with 7191562 segments
Show first segment. To see other segments just type its element number. E.g. D[7]

Attributes:  Dict("pga" => "0.379999995231628", "su_id" => "26107.000000000000000", "construc_1" => "{\"C99/LFINF+DNO/HBET:1,3\": 307.7814279, \"C99/LFINF+DNO/HBET:4,7\": 134.7351977, \"MUR+CL99/HBET:1,3\": 88.81801297}", "constructi" => "{\"C99/LFINF+DNO/HBET:1,3\": 3, \"C99/LFINF+DNO/HBET:4,7\": 1, \"MUR+CL99/HBET:1,3\": 1}", "osm_id" => "30962401.000000000000000", "E4" => "", "mean_FlowR" => "2.176470588235294", "std_FlowR" => "1.975958442049296")
BoundingBox: [338322.16972057807, 338343.0346088757, 3.0673524229467134e6, 3.0673803243858125e6]
PROJ: +proj=utm +zone=45 +datum=WGS84 +units=m +no_defs
WKT: PROJCS["WGS 84 / UTM zone 45N",
    GEOGCS["WGS 84",
        DATUM["WGS_1984",
            SPHEROID["WGS 84",6378137,298.257223563]],
        PRIMEM["Greenwich",0],
        UNIT["degree",0.0174532925199433]],
    PROJECTION["Transverse_Mercator"],
    PARAMETER["latitude_of_origin",0],
    PARAMETER["central_meridian",87],
    PARAMETER["scale_factor",0.9996],
    PARAMETER["false_easting",500000],
    PARAMETER["false_northing",0],
    UNIT["metre",1]]
5×2 GMTdataset{Float64, 2}
 Row │ X               Y         
     │ Float64         Float64   
─────┼───────────────────────────
   1 │      3.38322e5  3.06735e6
   2 │      3.38325e5  3.06738e6
   3 │ 338343.0        3.06738e6
   4 │      3.3834e5   3.06735e6
   5 │      3.38322e5  3.06735e6
Vector{GMTdataset} with 7191562 segments
Show first segment. To see other segments just type its element number. E.g. D[7]

I have two questions:
How can I access the information contained on the shapefile as a dataframe?
How can I create a GMTdataset from a dataframe with a geometry column?

Thank you for your help!

2

Quick answers as I won’t be able to follow up on this before the end of the afternoon.

  1. You’ll need to create the dataframe yourself. I would like to have more interfaces with DataFrames but I don’t want to make that package a GMT.jl dependency.

  2. You can create GMTdatsets with the function mat2ds (see the online help)

GMTdatasets have a lot of functionality. For example you can access columns by column name and attributes. Maybe you don’t need a dataframe. See also this example that uses dataframes.

And see also all fields of the GMTdatase type (it has a geometry field)

1 Like
3

Thanks -

you can access columns by column name and attributes

Great ! I would just need to know how to extract an attribute as array to add to a dataframe. Could you show me how to do that? I can’t seem to be able to figure it out with the documentation.

Also … that might be a different topic all together but I can’t change the position of the legend on the plot like in this example from pygmt:

fig.legend(position="jTL+o0.1c", box=True)

Thanks a lot !

4

Attributes can be used in filters but not directly to select columns. See this example that uses dataframes. The author refers to it as a header but probably due to the time of writing. Now the attributes field is a Dictionary. This other example also uses the attributes but it’s a bit more confusing specially if one do have the data files used in the example.

Regarding the labels position. Have you seen these examples?

1 Like
5

I am making good progress and I hope to fully integrate GMT.jl into my workflow.

I am left with 2 questions to avoid PyCall … I hope this is not redundant to your previous answer, I might have missed something.

You mentioned that it is possible to select a subset based on an attribute condition in a GMT dataset (e.g. pts[:, pts.A .= 999] for a Dataframe)

How can you do this in GMT.jl for a polygons dataset?

What is the fastest way to iterate through the rows in a polygon GMT dataset?

Thanks (again)

6

I’m not sure about your first question but GMTdatasets implement the AbstractArray and Tables interfaces so you can address to them as simple arrays. e.g.

for n = 1:size(D,1)
    println(D[n,:])                 # Print the D rows
end

Or simplly D[20:30, 60:90] (and they can be 3D too)

Edit: And if you want them to remain as a GMTdatasets use the function mat2ds(D::GMTdataset, inds), where inds is a matrix range, e.g. (20:30, 60:90)

7

Thanks I will try that.

yes, keeping it as a GMTdatasest was one of the problem I had :slight_smile: