[R-sig-Geo] Issues with a GDB file in R?

Thu May 3 13:19:32 CEST 2018

Hello Shaun,

Thanks a lot for replying and providing me alternative options.

Unfortunately, I can't try both options anymore as I ran out of ArcGIS due to I was using a university PC which is not available now (I installed an ArcGIS trial version there), but I'll try it later. I also failed on installing the sf package - I'll dig a bit more on it. Nevertheless, I already downloaded the SpatialLite version so I'll try the second option and I'll let you know how it goes.

Regards,
Roman.  

On 02/05/2018, 18:53, "Shaun Walbridge" <SWalbridge at esri.com> wrote:

    Hello Roman,

    A couple of suggested options: You can try and use the arcgisbinding package [1] to pull the data directly into R via ArcGIS, as you mentioned you have ArcGIS available [presumably on Windows or in a VM]. It will access the data directly, and let you create sp and sf objects out of Geodatabases with little work. A basic workflow looks like this:

    d <- arc.open("path/file.gdb/layer_name")
    df <- arc.select(d) # here, you can filter columns and attributes, see [2]
    # create an sf object
    df.sf <- arc.data2sf(df)
    # alternatively, create an sp object
    df.sp <- arc.data2sp(df)

    You can also write the results back to a geodatabase, or any other format that ArcGIS understands. Another option is trying the SpatiaLite version of the data at the URL you posted. You should be able to access this using R directly, provided your rgdal installation is correctly built to read spatialite databases. If it is, try the same process you mentioned using rgdal, but point it at the SpatialLite database instead. You could also use command-line OGR to convert the data, that has a few options like where clause filtering that aren't directly available via readOGR.

    If neither of these options work, let me know and I can convert the data for you into a format of your preference.

    Cheers,
    Shaun

    1. https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_R-2DArcGIS_r-2Dbridge-2Dinstall&d=DwIGaQ&c=n6-cguzQvX_tUIrZOS_4Og&r=yckcjR6k3nADiVsNiAhGwcZB--0A8DQgvLSJ27upmyk&m=dkecE8PWOn5sSrtSXqX2R-VwJuFH_In4BvpC5da4LJ0&s=nRO2IG9dlrdKYGMSp3rCf7nwJqRFQoIWnY5zSFrPOQM&e=
    2. https://urldefense.proofpoint.com/v2/url?u=https-3A__rdrr.io_github_R-2DArcGIS_r-2Dbridge_man_arc.select.html&d=DwIGaQ&c=n6-cguzQvX_tUIrZOS_4Og&r=yckcjR6k3nADiVsNiAhGwcZB--0A8DQgvLSJ27upmyk&m=dkecE8PWOn5sSrtSXqX2R-VwJuFH_In4BvpC5da4LJ0&s=89N-cLe5N3dhZ6MTsM2OAYwmDImBH88ZrVaA5Hos0n4&e=

    On 5/2/18, 1:34 PM, "R-sig-Geo on behalf of Aguirre Perez, Roman" <r-sig-geo-bounces at r-project.org on behalf of ra454 at exeter.ac.uk> wrote:

        Hi everyone,

        I’ve been struggling with a ESRI Geodatabase file (clc12_Version_18_5.gdb) which is a layer of land cover classes available on

        https://urldefense.proofpoint.com/v2/url?u=https-3A__land.copernicus.eu_pan-2Deuropean_corine-2Dland-2Dcover_clc-2D2012-3Ftab-3Ddownload&d=DwIGaQ&c=n6-cguzQvX_tUIrZOS_4Og&r=YFaRLkcUCdDkLrpTbNOUV9J1CwYBCTMwgm5tdQkRSm4&m=qIGI8rXMP_JTDhSFz8NjjQMAyNAnpnUGFRBxqfH4bPU&s=9LaO4HiB5C5ewiuN_IeSQJSgq2tl5_-oMECPChtZa_U&e=

        whose size is 2.61 GB once unzipped.

        My ultimate task is to overlay it with the last NUTS3 administrative boundaries shapefile (2013) available on

        https://urldefense.proofpoint.com/v2/url?u=http-3A__ec.europa.eu_eurostat_web_gisco_geodata_reference-2Ddata_administrative-2Dunits-2Dstatistical-2Dunits&d=DwIGaQ&c=n6-cguzQvX_tUIrZOS_4Og&r=YFaRLkcUCdDkLrpTbNOUV9J1CwYBCTMwgm5tdQkRSm4&m=qIGI8rXMP_JTDhSFz8NjjQMAyNAnpnUGFRBxqfH4bPU&s=VExYSoR8FY_vhZaPNJpoOx0ZCZNMU9hMTtlGJZj2joU&e=

        in order to compute the area covered by each class within each NUTS3 region.

        Despite the ease of friendly software for performing this task, haven’t been capable of doing it - GRASS didn’t load the file as the log reports problems with the polygons, QGIS shows a warning regarding a specific object and ArcGIS got frozen. I guess it’s because the PC I used doesn’t have enough capacity. Unfortunately, I don’t have access to a more powerful one.

        Anyway, I decided to try with R -after all, I’ll perform my analysis with it. So I started exploring this GDB with rgdal:

        ogrInfo(dsn="clc12_Version_18_5.gdb",layer="clc12_Version_18_5")

        Source: "/Users/Roman/Desktop/clc12gdb/clc12_Version_18_5.gdb",
        layer: "clc12_Version_18_5"
        Driver: OpenFileGDB;
        number of rows: 2370829
        Feature type: wkbPolygon with 3 dimensions
        Extent: (-2693292 -3086662) - (10037210 5440568)
        Null geometry IDs: 2156240
        CRS: +proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs
        Number of fields: 6
        name                      type  length  typeName
        1 code_12              4       3           String
        2 ID                         4       18         String
        3 Remark               4       20         String
        4 Area_Ha             2       0           Real
        5 Shape_Length   2       0           Real
        6 Shape_Area       2       0           Real

        Then I tried to load it typing

        clc<-readOGR(dsn="clc12_Version_18_5.gdb",layer="clc12_Version_18_5")

        After 3-5 minutes, it appears the following text:

        OGR data source with driver: OpenFileGDB
        Source: "/Users/Roman/Desktop/clc12shp/clc12_Version_18_5.gdb", layer: "clc12_Version_18_5"
        with 2370829 features
        It has 6 fields

        Unfortunately, after trying 5 times (each one took around 8 hours) I couldn’t get anything but the following messages:

        Warning messages:
        1: In readOGR(dsn = "clc12_Version_18_5.gdb", layer = "clc12_Version_18_5") :
        Dropping null geometries: 2156240
        2: In readOGR(dsn = "clc12_Version_18_5.gdb", layer = "clc12_Version_18_5") :
        Z-dimension discarded

        Could anyone advise me how to tackle it? I also would appreciate suggestions on how to work with geodatabases - it’s my first time I work with these kind of files so I don’t even know their structure.

        By the way, these are my computer and software specifications:

        MacBook Pro 2012
        Processor 2.6 GHz Intel Core i7
        Memory 16 GB DDR3
        R 3.5.0
        RStudio 1.1.447

        Best,
        Roman.