Skip to content

Commit

Permalink
fixes #2 by adding references to toc
Browse files Browse the repository at this point in the history
  • Loading branch information
@dljo
dljo committed May 27, 2024
1 parent 370fc41 commit 7701fe1
Showing 1 changed file with 61 additions and 54 deletions.
115 changes: 61 additions & 54 deletions doc/pgf.dox
View file
Original file line number Diff line number Diff line change
Expand Up @@ -16,11 +16,18 @@ to it.
\li \ref pgf_bak
\li \ref pgf_how
\li \ref pgf_prep
\li \ref pgf_ad
\li \ref pgf_qm
\li \ref pgf_vol
\li \ref pgf_cp
\li \ref pgf_comp
\li \ref pgf_site2d
\li \ref pgf_gm
\li \ref pgf_acrr
\li \ref pgf_gra
\li \ref pgf_wrwp
\li \ref pgf_scansun
\li \ref pgf_groovy_scripts
\li \ref pgf_cl
\li \ref pgf_scr
\li \ref pgf_io
Expand Down Expand Up @@ -117,7 +124,7 @@ and pressing the "Create" button, e.g. for the bRopo package:
\image html pgf_cqc.png
This is a little tricker than it first seems, because the PGF must
recognize the "Name" that you have entered. There are several quality controls
available and might differ depending on what product generator that is used and
available and might differ depending on what product generator that is used and
what components that has been installed.

At the time of this documents writing, the following quality controls are available:
Expand Down Expand Up @@ -165,7 +172,7 @@ for each type of scan strategy.
Here is an example of a volume route used with Swedish data.
\image html pgf_vol.png
As soon as the bottom scan is received, the rule will activate and will
monitor inbound scans from any of the selected sources.
monitor inbound scans from any of the selected sources.
As soon as the top scan is received for a given source, the job will
run. The timeout of 900 seconds (15 min) is very generous and can be
reduced. If the top scan hasn't been received before the timeout is
Expand All @@ -187,7 +194,7 @@ scans, quality controlling them if this has been selected, and then inject
the resulting volume into the node.

\section pgf_cp Defining geographic projections and Cartesian areas
BALTRAD software uses the
BALTRAD software uses the
<a href="http://http://trac.osgeo.org/proj/" target="_blank">PROJ.4</a>
projections library to navigate data. In the toolbox, there are tools
available to help you do this. Normally, we would refer directly to the
Expand Down Expand Up @@ -244,7 +251,7 @@ projection_resolution(m) = ground_resolution(m) x 1 / cos(lat_ts)
Our radar is close to 60 degrees north, so let's use this latitude of true
scale for the same of simplicity. If we want a real resolution on the
ground of one kilometer, the projection's resolution will be two
kilometers. Convenient. This is used when defining the Cartesian area:
kilometers. Convenient. This is used when defining the Cartesian area:
\verbatim
$ area_registry --make --identifier=sevar_gmaps \
--description="Google Maps area for Vara" --proj_id=gmaps --files=sevar.h5 \
Expand Down Expand Up @@ -307,7 +314,7 @@ available areas.

Prodpar is not used in the composite route as a selection filter. It only provides
the PGF with a value to use during the product generation. That means that there is
a limitation on the PPI method, only the lowest elevation will be used as a selection
a limitation on the PPI method, only the lowest elevation will be used as a selection
criteria if SCAN-based selection has been chosen.

The observant reader will probably recognize that the PVOL route above also
Expand All @@ -329,7 +336,7 @@ works. Everything's a composite ...

\section pgf_site2d Site 2D products
If you want to generate single site cartesian products you can use the Site 2D route
instead. It almost works like the composite route with a couple of exceptions.
instead. It almost works like the composite route with a couple of exceptions.

First, it triggers immediately when a source arrives matching the source-name and the prodpar. This
means that when specifying PPI, prodpar and SCAN-based. The exact angle as specified in prodpar will
Expand All @@ -353,7 +360,7 @@ you need to do manually before this service will work for you.

\subsection pgf_gm_ap Configuring your web server

First thing to do is preparing a file containing geographic area definitions
First thing to do is preparing a file containing geographic area definitions
to use with the Google Maps plugin. To generate areas definition file, do the
following:

Expand All @@ -368,10 +375,10 @@ The name of the file \c smhi-areas.xml might be a bit confusing, but it actually
contains different product definitions.
\endnote

Assuming that all went fine, a file called \c products.js will be created in
<tt>$prefix/rave_gmap/web</tt> directory. It contains some area definitions we
Assuming that all went fine, a file called \c products.js will be created in
<tt>$prefix/rave_gmap/web</tt> directory. It contains some area definitions we
have used while developing the system. They can be replaced with your new one,
or your new one can be added - refer to \ref pgf_gm_gm section to learn how to
or your new one can be added - refer to \ref pgf_gm_gm section to learn how to
do that.

Next, make sure your web server knows about this plugin. We have integrated
Expand All @@ -393,9 +400,9 @@ root# service httpd restart

\subsection pgf_gm_ap_alt Alternative web server configuration

In case your system doesn't come with pre-installed apache server, or for some
reason you need a custom setup, you might want to install it manually.
Here are the steps necessary to install apache2 web server under Ubuntu
In case your system doesn't come with pre-installed apache server, or for some
reason you need a custom setup, you might want to install it manually.
Here are the steps necessary to install apache2 web server under Ubuntu
operating system.

First, install the required packages:
Expand All @@ -404,24 +411,24 @@ First, install the required packages:
sudo apt-get install apache2
sudo apt-get install php5
sudo apt-get install libapache2-mod-php5
\endverbatim
\endverbatim

Then point the server to the right \p DocumentRoot directory. In order
Then point the server to the right \p DocumentRoot directory. In order
to achieve this, edit the following configuration file:

\verbatim
/etc/apache2/sites-available/default
\endverbatim

Set the \p DocumentRoot option so it points to your web data directory:
Set the \p DocumentRoot option so it points to your web data directory:

\verbatim
DocumentRoot /opt/baltrad/rave_gmap/web
\endverbatim

Save the file and restart your web server:

\verbatim
\verbatim
sudo service apache2 restart
\endverbatim

Expand All @@ -445,11 +452,11 @@ for Google Maps each time an input file for it is generated.
\subsection pgf_gm_gm Configuring the plugin itself
The final step is for you to configure the plugin itself, so that it
contains a selectable product menu item that matches the product you're
generating. We will use \c products.js file that you've generated in
generating. We will use \c products.js file that you've generated in
\ref pgf_gm_ap section. As was mentioned above, the file contains some default
area definitions. They can be replaced with your new one, or new one can be
added. Continuing with our example from Vara, a corresponding entry for it makes
use of the corner coordinates we determined earlier in \ref pgf_cp_area and the
area definitions. They can be replaced with your new one, or new one can be
added. Continuing with our example from Vara, a corresponding entry for it makes
use of the corner coordinates we determined earlier in \ref pgf_cp_area and the
radar's location:
\verbatim
//Product sevar_gmaps, Vara 1 km
Expand Down Expand Up @@ -484,7 +491,7 @@ job is triggered by a scheduler. So besides creating the route you will
have to add a scheduled job.

There are a number of things that the user has to be aware of when creating
the ACRR rule.
the ACRR rule.

- Object Type can be either IMAGE or COMP. This is the object type of the composites
that should be used in the accumulation. As a rule it should be COMP but you can
Expand All @@ -497,7 +504,7 @@ in the accumulation. If specifying 4 files per hour. The interval will be 15 min
- Acceptable loss is specified in percent (0-100) and describes how large loss that is accepted
before the accumulation is ignored.

\image html pgf_acrr.png
\image html pgf_acrr.png

After you have setup the ACRR route. It is time to setup the schedule. The easiest way to do this is
to use the provided scheduler in the user interface.
Expand Down Expand Up @@ -525,7 +532,7 @@ if __name__ == "__main__":
BaltradFrame.send_message(msg, "http://localhost:8080/BaltradDex")
\endverbatim

The ID is just for traceability so this can be whatever. The name is the name of the route.
The ID is just for traceability so this can be whatever. The name is the name of the route.
\b NOTE: Most of the routes does not support triggering.

\section pgf_gra Gauge Radar Adjustment (GRA)
Expand All @@ -542,7 +549,7 @@ by using the rave dom object \b rave_dom_db.rave_observation.
If you want to try to use the functionality provided by rave to import the synop data in FM12 format
you will be using the fm12_importer script in the rave bin folder. But first, you need to get the
fm12 messages somehow. One alternative is to subscribe them from GISC. The fm12_importer is using
inotify for monitoring one folder so the fm12 messages should be placed in that folder.
inotify for monitoring one folder so the fm12 messages should be placed in that folder.

Before you start the fm12_importer daemon you will also have to setup information about
the stations that are delivered in the fm12 messages. This can be done by importing a flatfile
Expand All @@ -554,10 +561,10 @@ Download the most recent file and import it with the following command
%> python bin/wmo_station --uri=postgresql://<dbuser>:<dbpwd>@<dbhost>/<dbname> --flatfile=<flatfile> import
\endverbatim

When that has been done, it's just to start the fm12 importer daemon.
When that has been done, it's just to start the fm12 importer daemon.

\verbatim
%> /opt/baltrad/rave/bin/fm12_importer --monitored=/gisc/fm12 --pidfile=/opt/baltrad/rave/etc/fm12_importer.pid
%> /opt/baltrad/rave/bin/fm12_importer --monitored=/gisc/fm12 --pidfile=/opt/baltrad/rave/etc/fm12_importer.pid
--logfile=/opt/baltrad/rave/etc/fm12_importer.log --catchup --janitor daemon
\endverbatim

Expand All @@ -568,7 +575,7 @@ As you can see, the command is quite self-explanatory. A couple of notes might b
have been processed
- daemon just says that the process should be started as a daemon.

\subscection rave_pgf_gra Creating the GRA route
\subsection rave_pgf_gra Creating the GRA route
The GRA route is very similar to the ACRR route in both how it behaves and
that it is triggered by a scheduler.

Expand All @@ -581,8 +588,8 @@ First term UTC is 6 and interval is 12, the periods will be 0600-1800, 1800-0600
Since the synops might arrive later than for example 0600, we have scheduled the coefficients
to be run one hour later, i.e. at 0700 and 1900.

If you are using the Rave PGF you should be aware that the coefficients can be generated for
any area but they will be the same for all GRA adjustments. Hence, just have one GRA route
If you are using the Rave PGF you should be aware that the coefficients can be generated for
any area but they will be the same for all GRA adjustments. Hence, just have one GRA route
that uses a decently sized area.

\b NOTE: It is not relevant to specify a too small interval, a good interval is probably
Expand All @@ -594,7 +601,7 @@ When you have setup and scheduled the GRA route to be executed the gra-coefficie
be generated. Since the number of radar-gauge observations will be quite small at any
given time we are using a merge period of a number of terms. This can be defined in the
variable MERGETERMS in rave/Lib/rave_defines.py. The current value assumes a 12-hour
interval and 20 terms giving a 10 day period for calculating the gra coefficients.
interval and 20 terms giving a 10 day period for calculating the gra coefficients.

\section pgf_wrwp Weather Radar Wind Profiles (WRWP)
The route for generating wind profiles is quite simple. It is triggered on volumes
Expand All @@ -606,7 +613,7 @@ The interval in this case is specifying how meters each section should be up to
maximum height. So in the case of max height = 12000 and interval = 200 the number
of sections will be 60.

\image html pgf_wrwp.png
\image html pgf_wrwp.png

\section pgf_scansun Scanning a polar volume for sun hits
We have added a route for scanning polar volumes for sun hits. The result of
Expand All @@ -615,7 +622,7 @@ each radar source.

You can find the resulting files in <install prefix>/rave/etc/scansun/.

\image html pgf_scansun.png
\image html pgf_scansun.png

\section pgf_groovy_scripts When you are missing a must have route
In some cases, the routes we have predefined are not enough for whatever reason.
Expand Down Expand Up @@ -668,14 +675,14 @@ class ScheduledComposite implements IScriptableRule {
private static String METHOD = "ppi"; // ppi, cappi, pcappi, pmax or max
private static String PRODPAR = "0.5";
private static String QUANTITY = "DBZH";

private static Logger logger = LogManager.getLogger(Ireland5MinuteComposite.class);
ExpressionFactory xpr;

ExpressionFactory xpr;
List sources = new ArrayList();
Expression dateTimeAttribute;
Formatter formatter;

Ireland5MinuteComposite() {
xpr = new ExpressionFactory();
formatter = new Formatter();
Expand All @@ -684,7 +691,7 @@ class ScheduledComposite implements IScriptableRule {
sources.add(xpr.literal(src))
}
}

@Override
public IBltMessage handle(IBltMessage message) {
if (message instanceof BltTriggerJobMessage) {
Expand All @@ -693,7 +700,7 @@ class ScheduledComposite implements IScriptableRule {
}
return null;
}

protected IBltMessage doHandle() {
DateTime now = getNow();
ArrayList<Expression> filters = new ArrayList<Expression>();
Expand All @@ -702,10 +709,10 @@ class ScheduledComposite implements IScriptableRule {
filters.add(xpr.in(xpr.attribute("_bdb/source_name"), xpr.list(sources)))
filters.add(xpr.ge(dateTimeAttribute, xpr.literal(getStartFromNow(now))))
filters.add(xpr.le(dateTimeAttribute, xpr.literal(getEndFromNow(now))))

FileQuery query = new FileQuery();
query.setFilter(xpr.and(filters));

ArrayList<String> files = new ArrayList<String>();
FileResult set = null;
try {
Expand All @@ -723,16 +730,16 @@ class ScheduledComposite implements IScriptableRule {
set.close();
}
}

BltGenerateMessage result = new BltGenerateMessage();
result.setAlgorithm("eu.baltrad.beast.GenerateComposite");
result.setFiles(files as String[]);
result.setFiles(files as String[]);
DateTime nominalTime = ManagerContext.getUtilities().createNominalTime(now, 5);
Date date = nominalTime.getDate();
Time time = nominalTime.getTime();
result.setArguments([
"--area="+AREA,
"--date="+new Formatter().format("%d%02d%02d",date.year(), date.month(), date.day()).toString(),
"--area="+AREA,
"--date="+new Formatter().format("%d%02d%02d",date.year(), date.month(), date.day()).toString(),
"--time="+new Formatter().format("%02d%02d%02d",time.hour(), time.minute(), time.second()).toString(),
"--selection="+SELECTION,
"--anomaly-qc=" + QC,
Expand All @@ -746,22 +753,22 @@ class ScheduledComposite implements IScriptableRule {
//"--qitotal_field=pl.imgw.quality.qi_total",
"--quantity=" + QUANTITY
] as String[])

for (String s : result.getArguments()) {
logger.info("Argument = " + s);
}

return result;
}

protected DateTime getNow() {
return DateTime.utcNow();
}

protected DateTime getStartFromNow(DateTime dt) {
return dt.add(new TimeDelta().addSeconds(-(7*60)));
}

protected DateTime getEndFromNow(DateTime dt) {
return dt.add(new TimeDelta().addSeconds(-10));
}
Expand All @@ -779,7 +786,7 @@ data contains a sun signature.
\verbatim
$ scansun polar_volume.h5
\endverbatim
An example from a Dutch radar should give e.g.
An example from a Dutch radar should give e.g.
\verbatim
#Date Time Elevatn Azimuth ElevSun AzimSun dBmMHzSun dBmStdd RelevSun Source
20110111 075022 0.30 126.50 -0.78 126.84 -113.31 0.6722 -0.19 RAD:NL51,NOD:nldhl
Expand All @@ -801,13 +808,13 @@ arguments and values are looked up from a file containing them is:
\verbatim
$ ropo -i input_volume.h5 -o output_volume.h5 --lookup=True
\endverbatim
An example result is displayed in \ref faq_ia. Run
An example result is displayed in \ref faq_ia. Run
\verbatim
$ ropo --help
\endverbatim
to get a full listing of options.

\section pgf_scr Data processing scripts
\section pgf_scr Data processing scripts
Scripting is an effective development method. Not only does it allow rapid
prototyping, it also reduces the bottlenecks in transforming a lab result
to an operational implementation. There are also other bottlenecks that are
Expand Down Expand Up @@ -920,7 +927,7 @@ And they have the following function that does something really simple:
\verbatim
int query(SCAN *myscan) {
int bytes_per_bin = 0;
printf("This scan's elevation angle is %2.1f degrees\n", myscan->elev);
printf("This scan's elevation angle is %2.1f degrees\n", myscan->elev);
printf("This scan has %ld rays\n", myscan->nazim);
printf("This scan has %ld bins per ray\n", myscan->nrang);
printf("The bin length is %3.1f m\n", myscan->rscale);
Expand Down Expand Up @@ -967,7 +974,7 @@ int map_ODIM_to_SCAN(PolarScan_t *odim_scan, SCAN *my_scan) {
memcpy(my_scan->data, PolarScanParam_getData(dbzh), bytes);
RAVE_OBJECT_RELEASE(dbzh);

/* Normally, we want to include enough checks to give different return
/* Normally, we want to include enough checks to give different return
codes if something goes wrong, but not in this simple case. */
return 0;
}
Expand Down

0 comments on commit 7701fe1

Please sign in to comment.