Geografic-Information-System/raster/r.in.xyz/r.in.xyz.html

<h2>DESCRIPTION</h2>

The <em>r.in.xyz</em> module will load and bin ungridded x,y,z ASCII data
into a new raster map. The user may choose from a variety of statistical
methods in creating the new raster. Gridded data provided as a stream of
x,y,z points may also be imported.

<p>
Please note that the current region extents and resolution are used for
the import. It is therefore recommended to first use the <b>-s</b>
flag to get the extents of the input points to be imported, then
adjust the current region accordingly, and only then proceed with the
actual import.

<p>
<em>r.in.xyz</em> is designed for processing massive point cloud datasets,
for example raw LIDAR or sidescan sonar swath data. It has been tested with
datasets as large as tens of billion of points (705GB in a single file).
 <!-- Doug Newcomb, US Fish & Wildlife Service -->

<p>
Available statistics for populating the raster are (<b>method</b>):
<p>
<blockquote>
<table>
<tr><td><em>n</em></td>        <td>number of points in cell</td></tr>
<tr><td><em>min</em></td>      <td>minimum value of points in cell</td></tr>
<tr><td><em>max</em></td>      <td>maximum value of points in cell</td></tr>
<tr><td><em>range</em></td>    <td>range of points in cell</td></tr>
<tr><td><em>sum</em></td>      <td>sum of points in cell</td></tr>
<tr><td><em>mean</em></td>     <td>average value of points in cell</td></tr>
<tr><td><em>stddev</em></td>   <td>standard deviation of points in cell</td></tr>
<tr><td><em>variance</em></td> <td>variance of points in cell</td></tr>
<tr><td><em>coeff_var</em></td><td>coefficient of variance of points in cell</td></tr>
<tr><td><em>median</em></td>   <td>median value of points in cell</td></tr>
<tr valign="baseline"><td><em>percentile</em>&nbsp;</td>
   <td>p<sup><i>th</i></sup> percentile of points in cell</td></tr>
<tr><td><em>skewness</em></td> <td>skewness of points in cell</td></tr>
<tr><td><em>trimmean</em></td> <td>trimmed mean of points in cell</td></tr>
</table>
</blockquote>

<ul>
<li><em>Variance</em> and derivatives use the biased estimator (n). [subject to change]
<li><em>Coefficient of variance</em> is given in percentage and defined as
<tt>(stddev/mean)*100</tt>.
</ul>

<p>
It is also possible to bin and store another data column (e.g. backscatter)
while simultaneously filtering and scaling both the data column values and
the z range.

<h2>NOTES</h2>

<h3>Gridded data</h3>

If data is known to be on a regular grid <em>r.in.xyz</em> can reconstruct
the map perfectly as long as some care is taken to set up the region
correctly and that the data's native map projection is used. A typical
method would involve determining the grid resolution either by examining
the data's associated documentation or by studying the text file. Next scan
the data with <em>r.in.xyz</em>'s <b>-s</b> (or <b>-g</b>) flag to find the
input data's bounds. GRASS uses the cell-center raster convention where
data points fall within the center of a cell, as opposed to the grid-node
convention. Therefore you will need to grow the region out by half a cell
in all directions beyond what the scan found in the file. After the region
bounds and resolution are set correctly with <em><a href="g.region.html">g.region</a></em>, run
<em>r.in.xyz</em> using the <i>n</i> method and verify that n=1 at all places.
<em><a href="r.univar.html">r.univar</a></em> can help. Once you are confident that the region exactly
matches the data proceed to run <em>r.in.xyz</em> using one of the <i>mean,
min, max</i>, or <i>median</i> methods. With n=1 throughout, the result
should be identical regardless of which of those methods are used.


<h3>Memory use</h3>

While the <b>input</b> file can be arbitrarily large, <em>r.in.xyz</em>
will use a large amount of system memory for large raster regions (10000x10000).
If the module refuses to start complaining that there isn't enough memory,
use the <b>percent</b> parameter to run the module in several passes.
In addition using a less precise map format (<tt>CELL</tt> [integer] or
<tt>FCELL</tt> [floating point]) will use less memory than a <tt>DCELL</tt>
[double precision floating point] <b>output</b> map. Methods such as <em>n,
min, max, sum</em> will also use less memory, while <em>stddev, variance,
and coeff_var</em> will use more.

The aggregate functions <em>median, percentile, skewness</em> and
<em>trimmed mean</em> will use even more memory and may not be appropriate
for use with arbitrarily large input files<!-- without a small value for percent= -->.
<!-- explained: memory use for regular stats will be based solely on region size,
 but for the aggregate fns it will also depend on the number of data points. (?) -->

<p>
The default map <b>type</b>=<tt>FCELL</tt> is intended as compromise between
preserving data precision and limiting system resource consumption.
If reading data from a <tt>stdin</tt> stream, the program can only run using
a single pass.


<h3>Setting region bounds and resolution</h3>

You can use the <b>-s</b> scan flag to find the extent of the input data
(and thus point density) before performing the full import. Use
<em><a href="g.region.html">g.region</a></em> to adjust the region bounds to match. The <b>-g</b> shell
style flag prints the extent suitable as parameters for <em><a href="g.region.html">g.region</a></em>.
A suitable resolution can be found by dividing the number of input points
by the area covered. e.g.

<div class="code"><pre>
wc -l inputfile.txt
g.region -p
# points_per_cell = n_points / (rows * cols)

g.region -e
# UTM location:
# points_per_sq_m = n_points / (ns_extent * ew_extent)

# Lat/Lon location:
# points_per_sq_m = n_points / (ns_extent * ew_extent*cos(lat) * (1852*60)^2)
</pre></div>

<p>
If you only intend to interpolate the data with <em><a href="r.to.vect.html">r.to.vect</a></em> and
<em><a href="v.surf.rst.html">v.surf.rst</a></em>, then there is little point to setting the region
resolution so fine that you only catch one data point per cell -- you might
as well use "<tt>v.in.ascii&nbsp;-zbt</tt>" directly.


<h3>Filtering</h3>

Points falling outside the current region will be skipped. This includes
points falling <em>exactly</em> on the southern region bound.
(to capture those adjust the region with "<tt>g.region s=s-0.000001</tt>";
see <em><a href="g.region.html">g.region</a></em>)
<p>Blank lines and comment lines starting with the hash symbol (<tt>#</tt>)
will be skipped.

<p>
The <b>zrange</b> parameter may be used for filtering the input data by
vertical extent. Example uses might include preparing multiple raster
sections to be combined into a 3D raster array with <em><a href="r.to.rast3.html">r.to.rast3</a></em>, or
for filtering outliers on relatively flat terrain.

<p>
In varied terrain the user may find that <em>min</em> maps make for a good
noise filter as most LIDAR noise is from premature hits. The <em>min</em> map
may also be useful to find the underlying topography in a forested or urban
environment if the cells are over sampled.

<p>
The user can use a combination of <em>r.in.xyz</em> <b>output</b> maps to create
custom filters. e.g. use <em><a href="r.mapcalc.html">r.mapcalc</a></em> to create a <tt>mean-(2*stddev)</tt>
map. [In this example the user may want to include a lower bound filter in
<em><a href="r.mapcalc.html">r.mapcalc</a></em> to remove highly variable points (small <em>n</em>) or run
<em><a href="r.neighbors.html">r.neighbors</a></em> to smooth the stddev map before further use.]


<h3>Alternate value column</h3>

The <b>value_column</b> parameter can be used in specialized cases when you
want to filter by z-range but bin and store another column's data. For
example if you wanted to look at backscatter values between 1000 and 1500
meters elevation. This is particularly useful when using <em>r.in.xyz</em>
to prepare depth slices for a 3D raster &mdash; the <b>zrange</b> option defines
the depth slice but the data values stored in the voxels describe an
additional dimension. As with the z column, a filtering range and scaling
factor may be applied.


<h3>Reprojection</h3>

If the raster map is to be reprojected, it may be more appropriate to reproject
the input points with <em><a href="m.proj.html">m.proj</a></em> or <em>cs2cs</em> before running
<em>r.in.xyz</em>.


<h3>Interpolation into a DEM</h3>

The vector engine's topographic abilities introduce a finite memory overhead
per vector point which will typically limit a vector map to approximately
3 million points (~ 1750^2 cells). If you want more, use the <em><a href="r.to.vect.html">r.to.vect</a></em>
<b>-b</b> flag to skip building topology. Without topology, however, all
you'll be able to do with the vector map is display with <em><a href="d.vect.html">d.vect</a></em> and
interpolate with <em><a href="v.surf.rst.html">v.surf.rst</a></em>.
Run <em><a href="r.univar.html">r.univar</a></em> on your raster map to check the number of non-NULL cells
and adjust bounds and/or resolution as needed before proceeding.

<p>
Typical commands to create a DEM using a regularized spline fit:
<div class="code"><pre>
r.univar lidar_min
r.to.vect -z type=point in=lidar_min out=lidar_min_pt
v.surf.rst in=lidar_min_pt elev=lidar_min.rst
</pre></div>

<h3>Import of x,y,string data</h3>

<em>r.in.xyz</em> is expecting numeric values as z column. In order to 
perform a occurrence count operation even on x,y data with non-numeric 
attribute(s), the data can be imported using either the x or y 
coordinate as a fake z column for <b>method</b>=<tt>n</tt> (count 
number of points per grid cell), the z values are ignored anyway.

<h2>EXAMPLES</h2>

<h3>Import of x,y,z ASCII into DEM</h3>

Sometimes elevation data are delivered as x,y,z ASCII files instead of a raster
matrix. The import procedure consists of a few steps: calculation of the
map extent, setting of the computational region accordingly with an additional
extension into all directions by half a raster cell in order to register the
elevation points at raster cell centers.
<p>
Note: if the z column is separated by several spaces from the coordinate columns,
it may be sufficient to adapt the <b>z</b> position value.

<!-- NC: preparation for this example
g.region raster=elevation res=15 -p 
# 
r.stats -1g elevation > elevation.xyz
-->

<div class="code"><pre>
# Important: observe the raster spacing from the ASCII file:
# ASCII file format (example):
# 630007.5 228492.5 141.99614
# 630022.5 228492.5 141.37904
# 630037.5 228492.5 142.29822
# 630052.5 228492.5 143.97987
# ...
# In this example the distance is 15m in x and y direction.

# detect extent, print result as g.region parameters
r.in.xyz input=elevation.xyz separator=space -s -g
# ... n=228492.5 s=215007.5 e=644992.5 w=630007.5 b=55.578793 t=156.32986

# set computational region, along with the actual raster resolution
# as defined by the point spacing in the ASCII file:
g.region n=228492.5 s=215007.5 e=644992.5 w=630007.5 res=15 -p

# now enlarge computational region by half a raster cell (here 7.5m) to
# store the points as cell centers:
g.region n=n+7.5 s=s-7.5 w=w-7.5 e=e+7.5 -p

# import XYZ ASCII file, with z values as raster cell values
r.in.xyz input=elevation.xyz separator=space method=mean output=myelev

# univariate statistics for verification of raster values
r.univar myelev
</pre></div>

<h3>Import of LiDAR data and DEM creation</h3>
Import the <a href="http://www.grassbook.org/ncexternal/index.html">Jockey's
Ridge, NC, LIDAR dataset</a> (compressed file "lidaratm2.txt.gz"), and process it
into a clean DEM:

<div class="code"><pre>
# scan and set region bounds
r.in.xyz -s -g separator="," in=lidaratm2.txt
g.region n=35.969493 s=35.949693 e=-75.620999 w=-75.639999
g.region res=0:00:00.075 -a

# create "n" map containing count of points per cell for checking density
r.in.xyz in=lidaratm2.txt out=lidar_n separator="," method=n zrange=-2,50

# check point density [rho = n_sum / (rows*cols)]
r.univar lidar_n
# create "min" map (elevation filtered for premature hits)
r.in.xyz in=lidaratm2.txt out=lidar_min separator="," method=min zrange=-2,50

# set computational region to area of interest
g.region n=35:57:56.25N s=35:57:13.575N w=75:38:23.7W e=75:37:15.675W

# check number of non-null cells (try and keep under a few million)
r.univar lidar_min

# convert to points 
r.to.vect -z type=point in=lidar_min out=lidar_min_pt

# interpolate using a regularized spline fit
v.surf.rst in=lidar_min_pt elev=lidar_min.rst

# set color scale to something interesting
r.colors lidar_min.rst rule=bcyr -n -e

# prepare a 1:1:1 scaled version for NVIZ visualization (for lat/lon input)
r.mapcalc "lidar_min.rst_scaled = lidar_min.rst / (1852*60)"
r.colors lidar_min.rst_scaled rule=bcyr -n -e
</pre></div>


<h2>TODO</h2>

<ul>
<li> Support for multiple map output from a single run.<br>
     <tt>method=string[,string,...] output=name[,name,...]</tt><br>
     This can be easily handled by a wrapper script, with the added
     benefit of it being very simple to parallelize that way.
</ul>


<h2>KNOWN ISSUES</h2>

<ul>
<li> "<tt>nan</tt>" can leak into <em>coeff_var</em> maps.
  <br>Cause unknown. Possible work-around: "<tt>r.null setnull=nan</tt>"
<!-- Another method:  r.mapcalc 'No_nan = if(map == map, map, null() )' -->
</ul>

If you encounter any problems (or solutions!) please contact the GRASS
Development Team.


<h2>SEE ALSO</h2>

<em>
<a href="g.region.html">g.region</a>,
<a href="m.proj.html">m.proj</a>,
<a href="r.fillnulls.html">r.fillnulls</a>,
<a href="r.in.ascii.html">r.in.ascii</a>,
<a href="r.in.lidar.html">r.in.lidar</a>,
<a href="r3.in.xyz.html">r3.in.xyz</a>,
<a href="r.mapcalc.html">r.mapcalc</a>,
<a href="r.neighbors.html">r.neighbors</a>,
<a href="r.out.xyz.html">r.out.xyz</a>,
<a href="r.to.rast3.html">r.to.rast3</a>,
<a href="r.to.vect.html">r.to.vect</a>,
<a href="r.univar.html">r.univar</a>,
<a href="v.in.ascii.html">v.in.ascii</a>,
<a href="v.surf.rst.html">v.surf.rst</a>
</em>
<p>
<em>
<a href="v.lidar.correction.html">v.lidar.correction</a>,
<a href="v.lidar.edgedetection.html">v.lidar.edgedetection</a>,
<a href="v.lidar.growing.html">v.lidar.growing</a>,
<a href="v.outlier.html">v.outlier</a>,
<a href="v.surf.bspline.html">v.surf.bspline</a>
</em>
<p>
<em><a href="http://www.ivarch.com/programs/pv.shtml">pv</a></em>
 - The UNIX pipe viewer utility

<p>
Overview: <a href="https://grasswiki.osgeo.org/wiki/Interpolation">Interpolation and Resampling</a> in GRASS GIS

<h2>AUTHORS</h2>

Hamish Bowman, Department of Marine Science, University of Otagom New Zealand
<br>
Extended by Volker Wichmann to support the aggregate functions
<i>median, percentile, skewness</i> and <i>trimmed mean</i>.

<p>
<i>Last changed: $Date$</i>