I ran into this problem as well today, and I had to install the GitHub development versions of ggplot2 and ggmap and restart R to get rid of this error:

    devtools::install_github(&quot;dkahle/ggmap&quot;)
    devtools::install_github(&quot;hadley/ggplot2&quot;)

Before that, I also reinstalled all of the packages mentioned here: https://github.com/thomasp85/ggraph/issues/10

Don&#39;t know if those reinstalls were necessary, as it was ultimately installing the GitHub version of ggmap that fixed the problem, but thought I&#39;d mention it just in case.

Note this problem appears to stem from the recent update to ggplot2 as discussed here: https://github.com/tidyverse/ggplot2/blob/master/NEWS.md#extensions

Reinstalling `ggmap` from source code can fix the problem.

    install.packages(&quot;ggmap&quot;, type = &quot;source&quot;)

The problem persisted for some calculations after following the steps above. Therefore I installed an older version of ggplot2.

    library(devtools)
    install_version(&quot;ggplot2&quot;, version = &quot;2.1.0&quot;, repos = &quot;http://cran.us.r-project.org&quot;)

I came across the same problem today and found that old saved ggmap images (as .RData) couldn&#39;t be read using the new ggplot2/ggmap versions (ggmap_2.6.1 &amp; ggplot2_2.2.0) resulting in a ggproto error. Admittedly this did drive me mad for quite sometime. However, when I reran the code and produced saved maps using the new versions of the ggmap and ggplot2 the images were totally readable. 

At this stage my suggestion is to rerun the code for any ggmap images you want to continue to access (preferred), or alternatively reinstall older versions of ggmap and ggplot2.

The error message is:

&gt; FATAL ERROR Uncaught Error: Cannot unpack array with string keys

I know I can simply run the method `fetch()` twice and pass the `[&#39;q&#39;]` and `[&#39;bind&#39;]`, but I am trying to get to grips with using the new `...` to unpack values. I want to pass the values like so:

    (string) SQL, (Array) Bind Values

But I think it tries to unpack the bind values array as well as the response array from the `fetch()` method. Is it possible to unpack this array?

It looks something like this:

    array(2) {
        [&quot;q&quot;] =&gt; string(7) &quot;example&quot;
        [&quot;bind&quot;] =&gt; array(1) {
            [&quot;example&quot;] =&gt; string(3) &quot;one&quot;
        }
    }

This is the whole code, in case you need to see how it all fits together:

    class ModelController {
        public static function execute($sql, $vals) {
            var_dump($vals);
        }
    }

    class ModelContainer {

        private $queries = [];

        public function add_model(Model $model, $name) {
            $this-&gt;queries[$name] = $model;
            return $this;
        }

        public function exec_all() {
            foreach($this-&gt;queries as $q) {
                ModelController::execute(...$q-&gt;fetch());
            }
        }

        public function exec($name) {

        }

    }

    class Model {

        private $sql;
        private $vals = [];


        public function set_q($statement) {
            $this-&gt;sql = $statement;
            return $this;
        }

        public function bind($vals = []) {
            $this-&gt;vals = $vals;
            return $this;
        }

        public function fetch() {
            return [&#39;q&#39; =&gt; (string)$this-&gt;sql,
                &#39;bind&#39; =&gt; $this-&gt;vals ];
        }
    }

    $m = new ModelContainer();
    $m-&gt;add_model((new Model)-&gt;set_q(&#39;example SQL content here&#39;)-&gt;bind([&#39;example&#39; =&gt; &#39;example value here&#39;]), &#39;one&#39;);
    $m-&gt;exec_all();







Cannot unpack array with string keys

I&#39;ve got this piece of code:

    $(document).ready(function () {
        $.getJSON(&#39;http://localhost:5000/&#39;, function (response) {
            console.log(response);
        });
    });

localhost:5000 is a `flask/python` script that returns a `json` like:

    {
      &quot;data&quot;: [
        0, 
        0, 
        0, 

And I&#39;m getting:

    $.getJSON is not a function TypeError: $.getJSON is not a function

Any tips where I can start untangling the whoolball?

Thanks!

Edit:

HTML:

    &lt;!DOCTYPE html&gt;
    &lt;html&gt;
    
    &lt;head&gt;
        &lt;script src=&quot;https://code.jquery.com/jquery-3.1.1.slim.min.js&quot;&gt;&lt;/script&gt;
        &lt;script src=&quot;lib/main.js&quot;&gt;&lt;/script&gt;
    &lt;/head&gt;
    
    &lt;body&gt;
    &lt;/body&gt;
    
    &lt;/html&gt;

`lib/main.js` is where the `document.ready` is located.

Thanks!

JQuery issue &quot;TypeError: $.getJSON is not a function&quot;

I&#39;m using ggmap, and got the error below:

    Error: GeomRasterAnn was built with an incompatible version of ggproto.
    Please reinstall the package that provides this extension.

I&#39;ve installed the latest version of both ggmap(2.6.1) and ggplot2(2.2.0), but still got the same error.


ggmap Error: GeomRasterAnn was built with an incompatible version of ggproto

I'm using ggmap, and got the error below:
<pre><code class="hljs language-javascript">Error: GeomRasterAnn was built with an incompatible version of ggproto.
Please reinstall the package that provides this extension.
</code></pre>
I've installed the latest version of both ggmap(2.6.1) and ggplot2(2.2.0), but still got the same error.

This simple code (and all my scripts from this morning) has started giving me an off center title in ggplot2:

    Ubuntu version: 16.04

    R studio version: Version 0.99.896

    R version: 3.3.2

    GGPLOT2 version: 2.2.0

I have freshly installed the above this morning to try and fix this...

    dat &lt;- data.frame(
    time = factor(c(&quot;Lunch&quot;,&quot;Dinner&quot;), levels=c(&quot;Lunch&quot;,&quot;Dinner&quot;)),
    total_bill = c(14.89, 17.23)
    )
    
    # Add title, narrower bars, fill color, and change axis labels
    ggplot(data=dat, aes(x=time, y=total_bill, fill=time)) + 
      geom_bar(colour=&quot;black&quot;, fill=&quot;#DD8888&quot;, width=.8, stat=&quot;identity&quot;) + 
      guides(fill=FALSE) +
      xlab(&quot;Time of day&quot;) + ylab(&quot;Total bill&quot;) +
      ggtitle(&quot;Average bill for 2 people&quot;)

[![enter image description here][1]][1]


  [1]: https://i.stack.imgur.com/v3zNm.png

Center Plot title in ggplot2

for some reason no link in my R markdowns (rmd) is formatted blue. knitting the simple rmd below to pdf is leaving the text color black. only when hovering over it does one realize that it&#39;s actually a link. knitting it to html will make the link blue. of course I can use a latex wrapper but I wonder why that is?

&gt; sessionInfo()
R version 3.3.0 (2016-05-03)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1
loaded via a namespace (and not attached):
 knitr_1.15

RStudio 1.0.44

    ---
    title: &quot;R Notebook&quot;
    output:
      pdf_document: default
      html_notebook: default
    ---

    ```{r, echo=F}
    # tex / pandoc options for pdf creation
    x &lt;- Sys.getenv(&quot;PATH&quot;)
    y &lt;- paste(x, &quot;E:\\miktex\\miktex\\bin&quot;, sep=&quot;;&quot;)
    Sys.setenv(PATH = y)
    ```

    [link](www.rstudio.com)

[![enter image description here][1]][1]
  [1]: https://i.stack.imgur.com/awos7.png

R markdown link is not formatted blue when knitted to pdf

How can I change the grey facet labels (A and B) into say red background with white text?
    

    library(data.table)
    A = data.table(x = 1:4, y = 1:4, z = c(&#39;A&#39;,&#39;A&#39;,&#39;B&#39;,&#39;B&#39;))
    ggplot(A) + geom_point(aes(x = x, y = y)) + facet_wrap(~z) + theme_bw()

[![enter image description here][1]][1]


  [1]: https://i.stack.imgur.com/OxFyG.png

Change facet label text and background colour

I am having difficulty grasping the essence of the `setDT()` function. As I read code on SO, I frequently come across the use of `setDT()` to create a data.table. Of course the use of `data.table()` is ubiquitous. I feel like I solidly comprehend the nature of `data.table()` yet the relevance of `setDT()` eludes me. `?setDT` tells me this:

&gt; `setDT` converts lists (both named and unnamed) and data.frames to data.tables _by reference_. 

as well as:

&gt;In `data.table` parlance, all `set*` functions change their input by reference. That is, no copy is made at all, other than temporary working memory, which is as large as one column.

So this makes me think I should only use `setDT()` to make a data.table, right? Is `setDT()` simply a list to data.table converter?  

    library(data.table)
    
    a &lt;- letters[c(19,20,1,3,11,15,22,5,18,6,12,15,23)]
    b &lt;- seq(1,41,pi)
    ab &lt;- data.frame(a,b)
    d &lt;- data.table(ab)
    e &lt;- setDT(ab)

    str(d)
    #Classes ‘data.table’ and &#39;data.frame&#39;:	13 obs. of  2 variables:
    # $ a: Factor w/ 12 levels &quot;a&quot;,&quot;c&quot;,&quot;e&quot;,&quot;f&quot;,..: 9 10 1 2 5 7 11 3 8 4 ...
    # $ b: num  1 4.14 7.28 10.42 13.57 ...
    # - attr(*, &quot;.internal.selfref&quot;)=&lt;externalptr&gt;
            
    str(e)
    #Classes ‘data.table’ and &#39;data.frame&#39;:	13 obs. of  2 variables:
    # $ a: Factor w/ 12 levels &quot;a&quot;,&quot;c&quot;,&quot;e&quot;,&quot;f&quot;,..: 9 10 1 2 5 7 11 3 8 4 ...
    # $ b: num  1 4.14 7.28 10.42 13.57 ...
    # - attr(*, &quot;.internal.selfref&quot;)=&lt;externalptr&gt;

Seemingly no difference in this instance. In another instance the difference is evident:

    ba &lt;- list(a,b)
    f &lt;- data.table(ba)
    g &lt;- setDT(ba)

    str(f)
    #Classes ‘data.table’ and &#39;data.frame&#39;:	2 obs. of  1 variable:
    # $ ba:List of 2
    #  ..$ : chr  &quot;s&quot; &quot;t&quot; &quot;a&quot; &quot;c&quot; ...
    #  ..$ : num  1 4.14 7.28 10.42 13.57 ...
    # - attr(*, &quot;.internal.selfref&quot;)=&lt;externalptr&gt;
    
    str(g)
    #Classes ‘data.table’ and &#39;data.frame&#39;:	13 obs. of  2 variables:
    # $ V1: chr  &quot;s&quot; &quot;t&quot; &quot;a&quot; &quot;c&quot; ...
    # $ V2: num  1 4.14 7.28 10.42 13.57 ...
    # - attr(*, &quot;.internal.selfref&quot;)=&lt;externalptr&gt;

When should I use `setDT()`? What makes `setDT()` relevant? Why not just make the original `data.table()` function capable of doing what `setDT()` is able to do?





 



 

        

 



When should I use setDT() instead of data.table() to create a data.table?

It appears that the raster package in R doesn&#39;t distinguish between positive and negative rotations of GeoTIFFs.  I have a feeling that it is because R is ignoring the negative signs in the rotation matrix.  I&#39;m not quite savvy enough to dig down into the `raster` source code to verify, but I did create a reproducible example that demonstrates the problem:
    
Read `R` logo and save as GeoTIFF.

    library(raster)
    b &lt;- brick(system.file(&quot;external/rlogo.grd&quot;, package=&quot;raster&quot;))
    proj4string(b) &lt;- crs(&quot;+init=epsg:32616&quot;)

    writeRaster(b, &quot;R.tif&quot;)

Add rotation to tiff with Python

    import sys
    from osgeo import gdal
    from osgeo import osr
    import numpy as np
    from math import *
    
    def array2TIFF(inputArray,gdalData,datatype,angle,noData,outputTIFF):
    #    this script takes a numpy array and saves it to a geotiff
    #    given a gdal.Dataset object describing the spatial atributes of the data set
    #    the array datatype (as a gdal object) and the name of the output raster, and rotation angle in degrees

    # get the file format driver, in this case the file will be saved as a GeoTIFF
      driver = gdal.GetDriverByName(&quot;GTIFF&quot;)

      #set the output raster properties
      tiff = driver.Create(outputTIFF,gdalData.RasterXSize,gdalData.RasterYSize,inputArray.shape[0],datatype)

      transform = []

      originX = gdalData.GetGeoTransform()[0]
      cellSizeX = gdalData.GetGeoTransform()[1]
      originY = gdalData.GetGeoTransform()[3]
      cellSizeY = gdalData.GetGeoTransform()[5]
      rotation = np.radians(angle)

      transform.append(originX)
      transform.append(cos(rotation) * cellSizeX)
      transform.append(sin(rotation) * cellSizeX)
      transform.append(originY)
      transform.append(-sin(rotation) * cellSizeY)
      transform.append(cos(rotation) * cellSizeY)

      transform = tuple(transform)

      #set the geotransofrm values which include corner coordinates and cell size
      #once again we can use the original geotransform data because nothing has been changed
      tiff.SetGeoTransform(transform)

      #next the Projection info is defined using the original data
      tiff.SetProjection(gdalData.GetProjection())

      #cycle through each band
      for band in range(inputArray.shape[0]):
          #the data is written to the first raster band in the image
          tiff.GetRasterBand(band+1).WriteArray(inputArray[band])

          #set no data value
          tiff.GetRasterBand(band+1).SetNoDataValue(0)

          #the file is written to the disk once the driver variables are deleted
      del tiff, driver

      inputTif = gdal.Open(&quot;R.tif&quot;)
      inputArray = inputTif.ReadAsArray()
    
      array2TIFF(inputArray,inputTif, gdal.GDT_Float64, -45, 0, &quot;R_neg45.tif&quot;)
      array2TIFF(inputArray,inputTif, gdal.GDT_Float64, 45, 0, &quot;R_pos45.tif&quot;)


Read in the rotated tiffs in `R`.

    c &lt;- brick(&quot;R_neg45.tif&quot;)
    plotRGB(c,1,2,3)
    d &lt;- brick(&quot;R_pos45.tif&quot;)
    plotRGB(d,1,2,3)
    
    &gt; c
    class       : RasterBrick 
    rotated     : TRUE
    dimensions  : 77, 101, 7777, 3  (nrow, ncol, ncell, nlayers)
    resolution  : 0.7071068, 0.7071068  (x, y)
    extent      : 0, 125.865, 22.55278, 148.4178  (xmin, xmax, ymin, ymax)
    coord. ref. : +proj=utm +zone=16 +datum=WGS84 +units=m +no_defs +ellps=WGS84 +towgs84=0,0,0 
    data source : /Users/erker/g/projects/uft/code/R_neg45.tif 
    names       : R_neg45.1, R_neg45.2, R_neg45.3 
    
    &gt; d
    class       : RasterBrick 
    rotated     : TRUE
    dimensions  : 77, 101, 7777, 3  (nrow, ncol, ncell, nlayers)
    resolution  : 0.7071068, 0.7071068  (x, y)
    extent      : 0, 125.865, 22.55278, 148.4178  (xmin, xmax, ymin, ymax)
    coord. ref. : +proj=utm +zone=16 +datum=WGS84 +units=m +no_defs +ellps=WGS84 +towgs84=0,0,0 
    data source : /Users/erker/g/projects/uft/code/R_pos45.tif 
    names       : R_pos45.1, R_pos45.2, R_pos45.3 

The plots are the same and note the equivalent extents.  However, `gdalinfo` tells a different story

    $ gdalinfo R_neg45.tif
    
    Driver: GTiff/GeoTIFF
    Files: R_neg45.tif
    Size is 101, 77
    Coordinate System is:
    PROJCS[&quot;WGS 84 / UTM zone 16N&quot;,
        GEOGCS[&quot;WGS 84&quot;,
            DATUM[&quot;WGS_1984&quot;,
                SPHEROID[&quot;WGS 84&quot;,6378137,298.257223563,
                    AUTHORITY[&quot;EPSG&quot;,&quot;7030&quot;]],
                AUTHORITY[&quot;EPSG&quot;,&quot;6326&quot;]],
            PRIMEM[&quot;Greenwich&quot;,0],
            UNIT[&quot;degree&quot;,0.0174532925199433],
            AUTHORITY[&quot;EPSG&quot;,&quot;4326&quot;]],
        PROJECTION[&quot;Transverse_Mercator&quot;],
        PARAMETER[&quot;latitude_of_origin&quot;,0],
        PARAMETER[&quot;central_meridian&quot;,-87],
        PARAMETER[&quot;scale_factor&quot;,0.9996],
        PARAMETER[&quot;false_easting&quot;,500000],
        PARAMETER[&quot;false_northing&quot;,0],
        UNIT[&quot;metre&quot;,1,
            AUTHORITY[&quot;EPSG&quot;,&quot;9001&quot;]],
        AUTHORITY[&quot;EPSG&quot;,&quot;32616&quot;]]
    GeoTransform =
      0, 0.7071067811865476, -0.7071067811865475
      77, -0.7071067811865475, -0.7071067811865476
    Metadata:
      AREA_OR_POINT=Area
    Image Structure Metadata:
      INTERLEAVE=PIXEL
    Corner Coordinates:
    Upper Left  (   0.0000000,  77.0000000) ( 91d29&#39;19.48&quot;W,  0d 0&#39; 2.50&quot;N)
    Lower Left  ( -54.4472222,  22.5527778) ( 91d29&#39;21.23&quot;W,  0d 0&#39; 0.73&quot;N)
    Upper Right (  71.4177849,   5.5822151) ( 91d29&#39;17.17&quot;W,  0d 0&#39; 0.18&quot;N)
    Lower Right (  16.9705627, -48.8650071) ( 91d29&#39;18.93&quot;W,  0d 0&#39; 1.59&quot;S)
    Center      (   8.4852814,  14.0674965) ( 91d29&#39;19.20&quot;W,  0d 0&#39; 0.46&quot;N)
    Band 1 Block=101x3 Type=Float64, ColorInterp=Gray
      NoData Value=0
    Band 2 Block=101x3 Type=Float64, ColorInterp=Undefined
      NoData Value=0
    Band 3 Block=101x3 Type=Float64, ColorInterp=Undefined
      NoData Value=0

    $ gdalinfo R_pos45.tif
    
    Driver: GTiff/GeoTIFF
    Files: R_pos45.tif
    Size is 101, 77
    Coordinate System is:
    PROJCS[&quot;WGS 84 / UTM zone 16N&quot;,
        GEOGCS[&quot;WGS 84&quot;,
            DATUM[&quot;WGS_1984&quot;,
                SPHEROID[&quot;WGS 84&quot;,6378137,298.257223563,
                    AUTHORITY[&quot;EPSG&quot;,&quot;7030&quot;]],
                AUTHORITY[&quot;EPSG&quot;,&quot;6326&quot;]],
            PRIMEM[&quot;Greenwich&quot;,0],
            UNIT[&quot;degree&quot;,0.0174532925199433],
            AUTHORITY[&quot;EPSG&quot;,&quot;4326&quot;]],
        PROJECTION[&quot;Transverse_Mercator&quot;],
        PARAMETER[&quot;latitude_of_origin&quot;,0],
        PARAMETER[&quot;central_meridian&quot;,-87],
        PARAMETER[&quot;scale_factor&quot;,0.9996],
        PARAMETER[&quot;false_easting&quot;,500000],
        PARAMETER[&quot;false_northing&quot;,0],
        UNIT[&quot;metre&quot;,1,
            AUTHORITY[&quot;EPSG&quot;,&quot;9001&quot;]],
        AUTHORITY[&quot;EPSG&quot;,&quot;32616&quot;]]
    GeoTransform =
      0, 0.7071067811865476, 0.7071067811865475
      77, 0.7071067811865475, -0.7071067811865476
    Metadata:
      AREA_OR_POINT=Area
    Image Structure Metadata:
      INTERLEAVE=PIXEL
    Corner Coordinates:
    Upper Left  (   0.0000000,  77.0000000) ( 91d29&#39;19.48&quot;W,  0d 0&#39; 2.50&quot;N)
    Lower Left  (  54.4472222,  22.5527778) ( 91d29&#39;17.72&quot;W,  0d 0&#39; 0.73&quot;N)
    Upper Right (      71.418,     148.418) ( 91d29&#39;17.17&quot;W,  0d 0&#39; 4.82&quot;N)
    Lower Right (     125.865,      93.971) ( 91d29&#39;15.42&quot;W,  0d 0&#39; 3.05&quot;N)
    Center      (  62.9325035,  85.4852814) ( 91d29&#39;17.45&quot;W,  0d 0&#39; 2.78&quot;N)
    Band 1 Block=101x3 Type=Float64, ColorInterp=Gray
      NoData Value=0
    Band 2 Block=101x3 Type=Float64, ColorInterp=Undefined
      NoData Value=0
    Band 3 Block=101x3 Type=Float64, ColorInterp=Undefined
      NoData Value=0

Is this a bug, or am I missing something? The `raster` package is incredibly powerful and useful, and I&#39;d rather help add more functionality than have to use other software to properly handle these (very annoyingly) rotated tiffs.  Thanks!  Also here&#39;s a R-sig-Geo [mailing post][1] related to rotated tiffs.


  [1]: https://stat.ethz.ch/pipermail/r-sig-geo/2011-May/011678.html

How to make R&#39;s &#39;raster&#39; package distinguish between positive and negative rotation matrices in GeoTIFFs?

Consider the following plot:

    library(ggplot2)

    p &lt;- ggplot(diamonds, 
                aes(x = carat, fill = cut)) +
      geom_density(position = &quot;stack&quot;) +
      facet_wrap(~ color)

[![annotated facet_wrap plot](https://i.stack.imgur.com/6OwZW.png)](https://i.stack.imgur.com/6OwZW.png)

The `facet_wrap` function wraps a sequence of faceted panels into a roughly rectangular display of `nrow` rows and `ncol` columns. However, depending on the data, the actual number of panels is often a few panels short of `nrow * ncol`, which leaves a chunk of wasted space in the plot.

If the plot includes legend(s), the situation is exacerbated, because now we have even more wasted space due to the legend, whether it&#39;s on the right (default legend position), or one of the other three directions.

To save space, I would like to shift the legend(s) into the space created by unfilled facets.

The following works as a space-saving measure, but the legend is anchored to a corner of the plot area, with potentially a lot of space left on one side, creating an imbalanced look:

    p +
      theme(legend.position = c(1, 0),
            legend.justification = c(1, 0))

[![legend anchored to a corner](https://i.stack.imgur.com/U2wPB.png)](https://i.stack.imgur.com/U2wPB.png)

Shifting a legend towards the centre of the blank space area by manually adjusting the `legend.position`/`legend.justification` values is a matter of trial and error, and difficult to scale if one has many faceted plots to work on.

In summary, I want a method that:

1. **Shifts the legend(s) of a faceted plot** into the space created due to empty facets.
2. Results in a **reasonably nice-looking** plot.
3. Is easily **automated** to handle many plots.

This is a recurring use case for me, and I&#39;ve decided to post it along with my working solution here in case anyone else finds it useful. I haven&#39;t seen this scenario *asked/answered* elsewhere on Stack Overflow. If anyone has, please leave a comment and I&#39;ll be happy to answer there instead or have this marked as a duplicate, as the case may be.

Content Type	Original Author	Original Content on Stackoverflow
Question	dingding	View Question on Stackoverflow
Solution 1 - R	jashu	View Answer on Stackoverflow
Solution 2 - R	Cyrille	View Answer on Stackoverflow
Solution 3 - R	Benjamin Ziepert	View Answer on Stackoverflow
Solution 4 - R	Nathan	View Answer on Stackoverflow

ggmap Error: GeomRasterAnn was built with an incompatible version of ggproto

R Problem Overview

R Solutions

Solution 1 - R

Solution 2 - R

Solution 3 - R

Solution 4 - R

JQuery issue "TypeError: $.getJSON is not a function"

Cannot unpack array with string keys

Attributions