[R] How to create gridded data

David L Carlson dc@r|@on @end|ng |rom t@mu@edu
Mon Mar 25 17:05:38 CET 2019


If the grid is not full you need to identify the missing cells. As an example we can remove rows 3 and 6 from DF1:

> DF1miss <- DF1[-c(3, 6), ]
> DF2miss <- xtabs(Precip~latitude+longitude, DF1miss)
> DF2miss
        longitude
latitude 110.5 111 111.5 112
    45.5   3.2 5.0   0.0 2.0
    46     6.1 0.0   7.8 5.5

The table command inserts 0 for empty cells, but for your data zero is a valid value so we need to identify the missing values and replace them with NA:

> DF2mod <- xtabs(~latitude+longitude, DF1miss) < 1
> DF2mod
        longitude
latitude 110.5   111 111.5   112
    45.5 FALSE FALSE  TRUE FALSE
    46   FALSE  TRUE FALSE FALSE

> DF2miss[DF2mod] <- NA
> DF2miss   # Print table with blanks for missing values:
        longitude
latitude 110.5 111 111.5 112
    45.5   3.2 5.0       2.0
    46     6.1       7.8 5.5
> print(DF2miss, na.print=NA)  # Print table with <NA> for missing values:
        longitude
latitude 110.5  111 111.5 112
    45.5   3.2  5.0  <NA> 2.0
    46     6.1 <NA>   7.8 5.5

----------------------------------------
David L Carlson
Department of Anthropology
Texas A&M University
College Station, TX 77843-4352





From: lily li <chocold12 using gmail.com> 
Sent: Monday, March 25, 2019 1:18 AM
To: David L Carlson <dcarlson using tamu.edu>
Cc: r-help <r-help using r-project.org>
Subject: Re: [R] How to create gridded data

Now I have new question about this post. If the grid coordinates in DF1 are not complete, i.e. there are missing coordinates, how to fill these with -99 in the exported DF2? Thanks.

On Thu, Nov 15, 2018 at 10:57 PM David L Carlson <mailto:dcarlson using tamu.edu> wrote:
It would depend on the format of the gridded data. Assuming it is a data frame like DF2 in my earlier answer, you just reverse the steps:

> DF2
     110.5 111 111.5 112
46     6.1 4.5   7.8 5.5
45.5   3.2 5.0   1.8 2.0

> DF3 <- data.frame(as.table(as.matrix(DF2)))
  Var1  Var2 Freq
1   46 110.5  6.1
2 45.5 110.5  3.2
3   46   111  4.5
4 45.5   111  5.0
5   46 111.5  7.8
6 45.5 111.5  1.8
7   46   112  5.5
8 45.5   112  2.0

But the latitude and longitude get converted to factors and we lose the column names:

> DF3 <- data.frame(as.table(as.matrix(DF2)))
> colnames(DF3) <- c("latitude", "longitude", "Precip")
> DF3$latitude <- as.numeric(as.character(DF3$latitude))
> DF3$longitude <- as.numeric(as.character(DF3$longitude))

----------------------------------------
David L Carlson
Department of Anthropology
Texas A&M University
College Station, TX 77843-4352

From: lily li <mailto:chocold12 using gmail.com> 
Sent: Tuesday, November 13, 2018 10:50 PM
To: David L Carlson <mailto:dcarlson using tamu.edu>
Cc: Sarah Goslee <mailto:sarah.goslee using gmail.com>; R mailing list <mailto:r-help using r-project.org>
Subject: Re: [R] How to create gridded data

Thanks, Sarah's answer helps the question. Now how to change the gridded data back to DF1 format? I don't know how to name the format, thanks.

On Tue, Nov 13, 2018 at 10:56 PM David L Carlson <mailto:mailto:dcarlson using tamu.edu> wrote:
Sarah's answer is probably better depending on what you want to do with the resulting data, but here's a way to go from your original DF1 to DF2:

> DF1 <- structure(list(latitude = c(45.5, 45.5, 45.5, 45.5, 46, 46, 46, 
+         46), longitude = c(110.5, 111, 111.5, 112, 110.5, 111, 111.5, 
+         112), Precip = c(3.2, 5, 1.8, 2, 6.1, 4.5, 7.8, 5.5)),
+         class = "data.frame", row.names = c(NA, -8L))
> 
# Convert to table with xtabs()
> DF2 <- xtabs(Precip~latitude+longitude, DF1)
> 

# Reverse the order of the latitudes
> DF2 <- DF2[rev(rownames(DF2)), ]
> DF2
        longitude
latitude 110.5 111 111.5 112
    46     6.1 4.5   7.8 5.5
    45.5   3.2 5.0   1.8 2.0

# Convert to a data frame
> DF2 <- as.data.frame.matrix(DF2)
> DF2
     110.5 111 111.5 112
46     6.1 4.5   7.8 5.5
45.5   3.2 5.0   1.8 2.0

----------------------------------------
David L Carlson
Department of Anthropology
Texas A&M University
College Station, TX 77843-4352


-----Original Message-----
From: R-help <mailto:mailto:r-help-bounces using r-project.org> On Behalf Of Sarah Goslee
Sent: Tuesday, November 13, 2018 8:16 AM
To: lily li <mailto:mailto:chocold12 using gmail.com>
Cc: r-help <mailto:mailto:r-help using r-project.org>
Subject: Re: [R] How to create gridded data

If you want an actual spatial dataset, the best place to ask is R-sig-geo

R has substantial capabilities for dealing with gridded spatial data,
including in the sp, raster, and sf packages.

Here's one approach, creating a SpatialGridDataFrame, which can be
exported in any standard raster format using the rgdal package.

DF2 <- DF1
coordinates(DF2) <- ~longitude + latitude
gridded(DF2) <- TRUE
fullgrid(DF2) <- TRUE

I recommend Roger Bivand's excellent book:
https://urldefense.proofpoint.com/v2/url?u=https-3A__www.springer.com_us_book_9781461476177&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=veMGHMCNZShld-KX-bIj4jRE_tP9ojUvB_Lqp0ieSdk&m=vZqNKoDe8N1TzBzeK12g2oa0cBS8VD6NDCs-hUhvt5o&s=B73PwZQrdKUmM1ML2Y5zjaEz7xqkHlzBDCrhluogK2U&e=

and there are abundant web tutorials.

Sarah
On Tue, Nov 13, 2018 at 2:22 AM lily li <mailto:mailto:chocold12 using gmail.com> wrote:
>
> Hi R users,
>
> I have a question about manipulating data. For example, I have DF1 as the
> following, how to transform it to a gridded dataset DF2? In DF2, each value
> Precip is an attribute of the corresponding grid cell. So DF2 is like a
> spatial surface, and can be imported to ArcGIS. Thanks for your help.
>
> DF1
> latitude   longitude   Precip
> 45.5           110.5         3.2
> 45.5           111            5.0
> 45.5           111.5         1.8
> 45.5           112            2.0
> 46              110.5         6.1
> 46              111            4.5
> 46              111.5         7.8
> 46              112            5.5
> ...
>
>
> DF2
> 6.1   4.5   7.8   5.5
> 3.2   5.0   1.8   2.0
> ...
>


-- 
Sarah Goslee (she/her)
https://urldefense.proofpoint.com/v2/url?u=http-3A__www.numberwright.com&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=veMGHMCNZShld-KX-bIj4jRE_tP9ojUvB_Lqp0ieSdk&m=vZqNKoDe8N1TzBzeK12g2oa0cBS8VD6NDCs-hUhvt5o&s=qSosThG59aeSFYzVFf1e-YQGbuBKVbvgVi1z9nFm884&e=

______________________________________________
mailto:mailto:R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
https://urldefense.proofpoint.com/v2/url?u=https-3A__stat.ethz.ch_mailman_listinfo_r-2Dhelp&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=veMGHMCNZShld-KX-bIj4jRE_tP9ojUvB_Lqp0ieSdk&m=vZqNKoDe8N1TzBzeK12g2oa0cBS8VD6NDCs-hUhvt5o&s=2pS9yFu5bpcRyCi1vX_OEDD2Ie8ZihvOQrkDQSNu8RM&e=
PLEASE do read the posting guide https://urldefense.proofpoint.com/v2/url?u=http-3A__www.R-2Dproject.org_posting-2Dguide.html&d=DwMFaQ&c=ODFT-G5SujMiGrKuoJJjVg&r=veMGHMCNZShld-KX-bIj4jRE_tP9ojUvB_Lqp0ieSdk&m=vZqNKoDe8N1TzBzeK12g2oa0cBS8VD6NDCs-hUhvt5o&s=1brzEOjZ4EUr_llqwyO274DfJfsOpRBI-pmd-hp0WAQ&e=
and provide commented, minimal, self-contained, reproducible code.


More information about the R-help mailing list