Product Set-up and Customization

EDEX Product Configuration

This automatic configuration is based on the afos2awips.txt pil/wmoid table. If there are errors in this table, then the products will not be correctly configured. To correct the templates after making changes to afos2awips.txt, you can rerun the template expansion scripts. Refer to the configureTextProduct User Guide for more details.

Some sites will have multiple entries in their afos2awips.txt table for a particular product, e.g., ZFP. The configuration script takes this into account and will automatically generate formatters for each entry found.

Quick Start Product Set Up

• Login to the GFE from a terminal window under a new user ID, e.g. TextTest. From a terminal window in the /awips2/cave directory, type ./cave.sh -u tester_id -component GFE. Login as a new user ID such as TextTest. You will use this ID to set up new products and test them. When you are satisfied with a product, you may move it to your SITE directory.
• Create Edit Areas. Create edit areas named "area1", "area2", and "area3". Some products are set up to run with these "defaultEditAreas" as noted in the table below. Other products run with a set of edit area Combinations which can be set up through the Formatter Launcher Dialog.
• Locate the product you wish to set up. From the GFE Main Menu, select "Define Text Products". Locate the "Standard" file for the product you wish to set up in the table below:

            NARRATIVE-TYPE PRODUCTS

• Open the Standard product file for viewing. Press the arrow to the left of the product name to display available configurations; then MB3 over the name of the standard file and choose, "Open".
• Create Data and Publish to Official Database. Locate the documentation section in the standard file for "Weather Elements Required." Create data for these weather elements and then, from the Products Menu, choose "Publish to Official."
• Run your product. From the "Products" menu, select "Formatter Launcher" The Formatter Launcher Dialog will appear listing the products. Select the name of the product, which brings up a new tab. If the product runs with Combinations of edit areas, a map will appear and you may use the Zone Combiner portion of the dialog to set up desired combinations. Choose "Run" to start the formatter.
• Make dialog selections. Some products will display a dialog for user input. Make appropriate selections and choose OK.
• Output will appear in the Formatter Launcher Dialog and you can modify, save, and/or transmit it.
• If you have any problems, refer to the Troubleshooting Guide.
• Review AreaDictionary. The "AreaDictionary" is located in the "TextUtilities" folder of the Localization perspective. This file is automatically generated and contains information about the zones, counties and cities for your site. This information is documented at the beginning of the file. Be sure and review the generated file and, if necessary, use MB3 "Open" to make changes at SITE level. Please note that any item listed as CONFIGURED is a generated item and will be overwritten during software upgrades.
• Set up Customized Edit Areas. For those products that use "defaultEditAreas", modify the Site Definition file with appropriate edit areas customized for your site.
  
• Run the Baseline, Regional, and Site Products.
• Deactivating Products: If products appear in the Formatter Launcher which your site does not want to use, you can deactivate them as follows: Open the Site Definition File OR Local product file and set the "displayName. Press MB3 over the Site Definition File OR Local file (see table above) and choose "Open". Set the "displayName" Definition entry. This is the name that will appear in the Formatter Launcher Dialog.
  

Hazard Products

Statements

Product Customization

Example: Customizing a Phrase

The "wind_withGusts_phrase" produces phrases such as:

WINDS NORTHEAST 10 TO 20 KNOTS WITH GUSTS TO AROUND 70 KNOTS VEERING SOUTHEAST 5 TO 15 KNOTS WITH GUSTS AROUND 25 KNOTS IN THE EARLY MORNING.

When we find this phrase in the Wind and WindGust Text Rules section, we see that it can be customized in various ways. If we want to change the descriptor from "WINDS" to "WIND", we need to override the phrase_descriptor_dict for "Wind". First, we look to see if it is already overridden in the Standard File, AreaFcst, which we open in the Text Products folder of the Define Text Products dialog. We can issue a "find" command to see if it is there. It is not, so next we look in the ConfigVariables utility from the Text Utilities folder and issue a "find" command. NOTE: The ConfigVariables Utility is "Read-Only", so you cannot change the method in that file. You MUST copy it into your Overrides file, and in that location, you can change the value which will override or supercede the value found in the library module.

We notice that the phrase_descriptor_dict is very lengthy. We do not have to repeat the entire method. Instead, we can use the dictionary as a starting point and simply change any entries that are relevant to us as follows:

def phrase_descriptor_dict(self, tree, node):
        dict = TextRules.TextRules.phrase_descriptor_dict(self, tree, node)
        dict["Wind"] = "wind"
        return dict

If we want to report knots as "KTS" instead of "KNOTS", we will override the units_descriptor_dict for "kt" as follows in the "Overriding Thresholds and Variables" section of our Overrides file:

  def units_descriptor_dict(self, tree, node):
        dict = TextRules.TextRules.units_descriptor_dict(self, tree, node)
        unitsDict = dict["units"]
        unitsDict["kt"] = "kts"
        return dict

       def gust_wind_difference_threshold(self, tree, node):
             # Difference between gust and max wind below which gusts are not mentioned
             # Units are in mph or knots depending on the current product
             return 15

Narrative Definition and Configurable Issuance -- issuance_list

Narrative products are composed of a "narrative" which contains a series of "component" phrase products. This is used to construct a "narrative" type of product of the form:

        {
            "type": "narrative",
            "narrativeDef":  [(<componentName>, <number of hours>), ...],
            "methodList": [self.assembleChildWords"],
            # optional:
           "priorPeriod": <number of hours>,   # Number of hours for sampling prior to forecast. Used for
                                                                              # first period trends
        }

For example:

       {
            "type": "narrative",
            "narrativeDef":  [("Period_1", 12), ("Period_2_3", 12), ("Period_2_3, 12)],
            "methodList": [self.assembleChildWords"],
            # optional:
           "priorPeriod": 24,   # Number of hours for sampling prior to forecast. Used for
                                              # first period trends
        }

Note that you may define a "Phantom" period during which no phrases will be generated.  It will act as a placeholder advancing your time periods:

       "narrativeDef": {
                ("Period", "period1"),
                ("Period", 12), ("Period", 12),
                ("LaterPeriod", 12), ("LaterPeriod", 12), ("LaterPeriod", 12),
                ("Phantom", 6), ("Extended", 24), ("Extended", 24), ("Extended", 24)
          }

You can customize the issuance times and the series of components by overriding the _issuance_list method, which returns information about the issuance times you want for your product.  The issuance_list, is found in the Standard file for narrative-type products.

 For each issuance time we must give the following information:

• description -- text to appear in the startUp dialog for the product.
• startHour, endHour -- start and end hours for the first period.  These times are relative to self.DAY() and self.NIGHT() which default to 6 and 18, respectively.  The startHour or endHour can be specified as "issuanceHour".
• expirationHour -- hour when the product expires in local time (it is converted to UGC for the header)
• period1 Label  -- the label for the first period. e.g. ".TODAY...", ".REST OF TODAY..."
• period1 lateNight phrase -- phrase to use if the hours of 3am to 6am must be qualified e.g. "Partly cloudy in the early morning."
• period1 lateDay phrase -- phrase to use if the hours of 3pm to 6pm must be qualified e.g. "Partly cloudy in the early evening."
• todayFlag -- if 1, "TODAY" and "TONIGHT" phrasing will be used in subsequent periods, otherwise, weekday wording will apply.
• narrative  definition -- component and time period pairs

   def _10_503_issuance_list(self, argDict):
        narrativeDefAM = [
            ("Period_1", "period1"),           
            ("Period_2_3", 12), ("Period_2_3", 12),
            ("Period_4_5", 12), ("Period_4_5", 12),
            ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12),
            ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12),
            ]
        narrativeDefPM = [
            ("Period_1", "period1"),
            ("Period_2_3", 12), ("Period_2_3", 12),
            ("Period_4_5", 12), ("Period_4_5", 12),
            ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12),
            ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12), ("Period_6_14", 12),
            ("Period_6_14", 12),
            ]
       
        return [
            ("Morning", self.DAY(), self.NIGHT(), self.NIGHT(),
             ".TODAY...", "early in the morning", "late in the afternoon",
             1, narrativeDefAM),
            ("Morning with Pre-1st Period", self.DAY()-2, self.NIGHT(), self.NIGHT(),
             ".TODAY...", "early in the morning", "late in the afternoon",
             1, narrativeDefAM),
            ("Morning Update", "issuanceHour", self.NIGHT(), self.NIGHT(),
             ".REST OF TODAY...", "early in the morning", "late in the afternoon",
             1, narrativeDefAM),
            ("Afternoon Update", "issuanceHour", self.NIGHT(), self.NIGHT(),
             ".REST OF TODAY...", "early in the morning","late in the afternoon",
             1, narrativeDefAM),
            #  End times are tomorrow:
            ("Afternoon", self.NIGHT(), 24 + self.DAY(), 24 + self.DAY(),
             ".TONIGHT...", "late in the night", "early in the evening",
             1, narrativeDefPM),
            ("Afternoon with Pre-1st Period", self.NIGHT()-2, 24 + self.DAY(), 24 + self.DAY(),
             ".TONIGHT...", "late in the night", "early in the evening",
             1, narrativeDefPM),
            ("Evening Update", "issuanceHour", 24 + self.DAY(), 24 + self.DAY(),
             ".REST OF TONIGHT...", "early in the morning","early in the evening",
             1, narrativeDefPM),
            # For the early morning update, this produces:
            # REST OF TONIGHT:
            # MONDAY
            # MONDAY NIGHT
            ("Early Morning Update", "issuanceHour", self.DAY(), self.DAY(),
             ".REST OF TONIGHT...", "early in the morning","late in the afternoon",
             0, narrativeDefPM),
            # Alternative
            # For the early morning update, this produces:
            # EARLY THIS MORNING:
            # TODAY
            # TONIGHT
            #("Evening Update", "issuanceHour", 24 + self.DAY(), 24 + self.DAY(),
            # ".REST OF TONIGHT...", "late in the night", "early in the evening",
            # 1, narrativeDefPM),
            #("Early Morning Update", "issuanceHour", self.DAY(), self.DAY(),
            # ".EARLY THIS MORNING...", "early in the morning", "late in the afternoon",
            # 1, narrativeDefPM),
            ]

NOTE: Any variables that begin with a "self._" should be encased in a "try" block. This is because the "_issuance_list" is accessed and examined once BEFORE the product is actually run in order to set up the input dialog. During this first examination, the variables are not yet initialized. As an example, here is the issuance_list from the FWF:

    def _issuance_list(self, argDict):
        narrativeDefAM = [
            ("FirePeriod", "period1"),
            ("FirePeriod", 12), ("FirePeriod", 12),
           ]
        narrativeDefPM = [
            ("FirePeriod", "period1"),
            ("FirePeriod", 12), ("FirePeriod", 12), ("FirePeriod", 12)
            ]
        extended = [
            ("FireExtendedShortTerm", 24), ("FireExtendedShortTerm", 24),
            ("FireExtendedShortTerm", 24),
            ("FireExtended", 24), ("FireExtended", 24), 
            ]
        try:
            if self._individualExtended == 1:
                if self._extendedLabel == 1:
                    narrativeDefAM.append(("ExtendedLabel",0))
                    narrativeDefPM.append(("ExtendedLabel",0))
                narrativeDefAM = narrativeDefAM + extended
            narrativeDefPM = narrativeDefPM + extended
        except:
            pass
       
        return [
            ("Morning", self.DAY(), self.NIGHT(), self.NIGHT(),
             ".TODAY...", "early in the morning", "late in the afternoon",
             1, narrativeDefAM),
            ("Morning Update", "issuanceHour", self.NIGHT(), self.NIGHT(),
             ".REST OF TODAY...", "early in the morning", "late in the afternoon",
             1, narrativeDefAM),
            ("Afternoon Update", "issuanceHour", self.NIGHT(), self.NIGHT(),
             ".REST OF TODAY...", "early in the morning","late in the afternoon",
             1, narrativeDefAM),
            #  End times are tomorrow:
            ("Afternoon", self.NIGHT(), 24 + self.DAY(), 24 + self.DAY(),
             ".TONIGHT...", "late in the night", "early in the evening",
             1, narrativeDefPM),
            ("Evening Update", "issuanceHour", 24 + self.DAY(), 24 + self.DAY(),
             ".REST OF TONIGHT...", "late in the night","early in the evening",
             1, narrativeDefPM),
            # For the early morning update, this produces:
            # REST OF TONIGHT:
            # MONDAY
            # MONDAY NIGHT
            ("Early Morning Update", "issuanceHour", self.DAY(), self.DAY(),
             ".REST OF TONIGHT...", "early in the morning","late in the afternoon",
             0, narrativeDefPM),
            # Alternative
            # For the early morning update, this produces:
            # EARLY THIS MORNING:
            # TODAY
            # TONIGHT
            #("Evening Update", "issuanceHour", 24 + self.DAY(), 24 + self.DAY(),
            # ".REST OF TONIGHT...", "late in the night", "early in the evening",
            # 1, narrativeDefPM),
            #("Early Morning Update", "issuanceHour", self.DAY(), self.DAY(),
            # ".EARLY THIS MORNING...", "early in the morning", "late in the afternoon",
            # 1, narrativeDefPM),
            ]
 

Post-Processing Narrative Phrases

Although, it is not encouraged, there will be times that you will want to post-process phrases to clean up undesirable wording.

A "postProcessPhrase" hook is available  for most phrases. This  allows you to do string replacement and/or abbreviations. Here is the default method from the PhraseBuilder module. You may copy it into your Overrides file and add replacements:

    def postProcessPhrase(self, tree, node):
        words = node.get("words")
        if words is not None:
            words = words.replace("rain showers and thunderstorms", "showers and thunderstorms")
            return self.setWords(node, words)

WFO Site Information