diff --git a/docs/Contributors_Guide/add_use_case.rst b/docs/Contributors_Guide/add_use_case.rst index c7ac6560cd..cc884963eb 100644 --- a/docs/Contributors_Guide/add_use_case.rst +++ b/docs/Contributors_Guide/add_use_case.rst @@ -108,7 +108,7 @@ configuration file name excluding the .conf suffix. .. figure:: figure/model_applications_example.png .. figure:: figure/model_applications_subdir.png - + Use Case Rules -------------- diff --git a/docs/Users_Guide/getting_started.rst b/docs/Users_Guide/getting_started.rst new file mode 100644 index 0000000000..3aada6c852 --- /dev/null +++ b/docs/Users_Guide/getting_started.rst @@ -0,0 +1,379 @@ +*************** +Getting Started +*************** + +This chapter reviews some important things to consider before starting +a METplus project. + +Questions to Consider +===================== + +Questions to ask when applying METplus to a new testing and evaluation project +------------------------------------------------------------------------------ + +If a user is new to the concept of developing a verification or +evaluations system, there are questions that should be considered to help +determine which tools to use and how to set up METplus. + +* First and foremost, what are the questions that need to be answered + with this testing and evaluation project? + +* What type of forecasts and type of observations will be used and how + can/should they be matched? + + * What attributes of the forecast should be evaluated? + + * What is the standard for comparison that provides a reference level + of skill (e.g., persistence, climatology, reference model)? + + * What is the geographic location of the model data being evaluated? + Are there specific areas of interest for the evaluation? + + * What domain should be used to evaluate on: The model domain, + observation domain (if gridded), or some other domain? + + * What is the evaluation time period? + Retrospective with specific dates? + Ongoing, near-real-time evaluation? + Or both retrospective and real-time? + +* How should the testing and evaluation project be broken down into + METplus Use Cases? One large one or multiple smaller ones? + + * How will METplus be run? Manually? Scheduled, through cron? + Automated via a workflow manager (e.g. Rocoto, EC-Flow, Rose-Cylc)? + + * Where will METplus be run? Local machine, project machine, + HPC system, in the cloud (e.g. AWS)? Serial runs or parallelized? + +This section will provide some guidance on how to use METplus based on +the answers. + +* What type of forecasts and type of observations will be used? Will they be + gridded or point-based? The METplus tools that can be used will vary + depending on the answer. Here’s a matrix to help: + +.. role:: raw-html(raw) + :format: html + +.. list-table:: METplus Tools Decision Matrix + :widths: auto + :header-rows: 1 + + * - + - Gridded Forecast + - Point Forecast + * - **Gridded** :raw-html:`
` + **Observation/Analysis** + - Gen-Ens-Prod :raw-html:`
` + PCP-Combine :raw-html:`
` + Ensemble-Stat :raw-html:`
` + Grid-Stat :raw-html:`
` + Wavelet-Stat :raw-html:`
` + MODE :raw-html:`
` + MTD :raw-html:`
` + Series-Analysis :raw-html:`
` + Grid-Diag :raw-html:`
` + TC-Gen :raw-html:`
` + TC-RMW :raw-html:`
` + - Point-Stat (run with the Analyses :raw-html:`
` + as the “forecast” and Point Forecast :raw-html:`
` + as the “observation”) + * - **Point Observations** + - Point-Stat :raw-html:`
` + Ensemble-Stat + - Stat-Analysis (run by passing in MPR records) :raw-html:`
` + TC-Pairs :raw-html:`
` + TC-Gen :raw-html:`
` + TC-Stat + + +* What attributes of the forecast should be evaluated? + + * This refers to defining the forecast fields to be evaluated, + as well as the forecast characteristics such as bias, reliability, + resolution, and prediction of events. It also means understanding + the nature of the forecast and observations. + +Examples of the nature of fields to be evaluated +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +* Continuous fields – the values change at the decimal level. + +* Categorical fields – the values change incrementally most + likely as integers or categories. Continuous fields can also be + turned into categorical fields via applying thresholds. + +* Probability fields – the values represent the probability or + likelihood of an event occurring, usually represented by thresholds. + +* Ensemble fields – are made up of multiple predictions either from + the same modeling system or multiple systems. + +Definitions of statistics categories associated with each type of field: + +* Continuous statistics - measures how the values of the forecasts + differ from the values of the observations. + + * METplus line types: SL1L2, SAL1L2, VL1L2, VAL1L2, CNT, VCNT. + + * METplus tools: + +* Categorical statistics - measures how well the forecast captures events. + + * METplus line types: FHO, CTC, CTS, MCTC, MCTS, ECLV, TC stats, + ExtraTC stats, TC-Gen stats. + +* Probability statistics - measures attributes such as reliability, + resolution, sharpness, and uncertainty. + + * METplus line types: PCT, PSTD, PJC, PRC. + +* Ensemble statistics - measures attributes as the relationship between + rank of observation and members, spread of ensemble member solutions + and continuous measures of skill. + +Additional verification and diagnostic approaches that can be helpful +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +* Geographical methods demonstrate where the error occurs geographically. + + * METplus methods: Series-Analysis tool. + + * METplus line types: Most Grid-Stat and Point-Stat line types. + +* Object Based measures the location error of the forecast and how the + total error break down into variety of descriptive attributes. + + * METplus methods: MODE, MTD, MvMODE, Grid-Stat Distance Maps. + + * METplus line types: MODE object attribute files, MODE CTS, MTD object + attribute files, MTD CTS, Grid-Stat DMAP. + +* Neighborhood relaxes the requirement for an exact match by evaluating + forecasts in the local neighborhood of the observations. + + * METplus methods: Grid-Stat Neighborhood, Point-Stat HiRA, Ensemble-Stat + HiRA. + + * METplus line types: NBRCTC, NBRCTS, NBRCNT, ECNT, ORANK, RPS. + +* Domain Decomposition and Transform applies a transform to a given field + to identify errors on different spatial scales: + + * METplus methods: Grid-Stat Fourier Decomposition; Wavelet-Stat tool, + TC-RMW tool. + + * METplus line types: Grid-Stat SL1L2, SAL1L2, VL1L2, VAL1L2, CNT, VCNT; + Wavelet Stat: ISC, RMW output file. + +* Feature Relative identifies systematic errors associated with a group + of case studies. + + * METplus methods: Feature Relative Use Cases. + +* Relationship between two fields: generates a joint PDF between two fields. + + * METplus methods: Grid-Diag tool. + +* Subseasonal-to-Seasonal Diagnostics compute indices to establish the + ability of the model to predict S2S drivers. + + * METplus methods: S2S Use Cases. + +What is the standard for comparison that provides a reference level of skill (e.g., persistence, climatology, reference model)? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Climatologies or Reference models may be passed into METplus using the +following configuration options: + +* {MET TOOL}_CLIMO_MEAN + +* {MET TOOL}_CLIMO_STDEV + +* {MET TOOL}_CLIMO_CDF + +This can be found in Grid-Stat, Point-Stat, Gen-Ens-Prod, Series-Analysis, +and Ensemble-Stat tools. + +What is the geographic location of the model data being evaluated? Are there specific areas of interest for the evaluation? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Masking regions are what METplus uses to define verification areas of +interest. These can be defined prior to running tools using the +Gen-Vx-Mask tool, or during run-time using the METPLUS_MASK_DICT options. + +What domain should be used for evaluation: The model domain, observation domain (if gridded), or some other domain? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +The decision to evaluate on model or observation/analysis domain is +user-specific but the user may want to consider the following: + +* Regridding to the courser domain will smooth high resolution information + that may be important but smoother forecasts tend to score better. + +* Regridding to a finer domain essentially adds in additional information + that is not real. + +* One way to avoid the interpolation debate is to regrid both to a third + grid. + +Regridding in METplus can be completed using the Regrid-Data-Plane tool if +the fields will be used more than once. + +Regridding can also be done on the fly using the {Tool}_REGRID_TO_GRID. +All grid-to-grid verification tools have the regridding capability in it. + +What is the evaluation time period? Retrospective with specific dates? Ongoing, near-real-time evaluation? Or both retrospective and realtime? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Basically, running retrospectively means that the observations/analyses are +already available on disk and running in realtime is when the system needs +to wait for the observations to be available on the system. + +In METplus, the LOOP_BY configuration can be used. + +LOOP_BY = VALID or REALTIME to have METplus proceed through the data based +on Valid Time. + +LOOP_BY = INIT or RETRO to have METplus proceed through the data based +on Initialization Time. + +How should the testing and evaluation project be broken down into METplus Use Cases? One large one or multiple smaller ones? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +A single use case is typically appropriate for a given evaluation so that +all of the information is found in one configuration file. However, users +may want to evaluate different combinations of models and observations. +For example, they may want to compare forecastA with observationA, +forecastA with observationB, forecastB with observationA, forecastB with +observationB, etc. In this case, separate METplus configuration files can +be created with information specific to each forecast or observation. +Another configuration file can be used to control settings common to each +evaluation, such as timing information and the process list. The METplus +wrappers can be called with each desired combination. + +.. code-block:: ini + + run_metplus.py forecastA.conf observationA.conf use_case_name.conf + run_metplus.py forecastA.conf observationB.conf use_case_name.conf + run_metplus.py forecastB.conf observationA.conf use_case_name.conf + run_metplus.py forecastB.conf observationB.conf use_case_name.conf + +It is also worth considering the +`Use Case Rules. `_ +A case may be +affected by the size of the data, the length of time to run and other factors. + +How will METplus be run? Manually? Scheduled through cron? Automated via a workflow manger (e.g. Rocoto, EC-Flow, Rose-Cylc)? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + * If run manually, this can be done. + + * If scheduled through cron, a bash or csh script can be written to + set up environment variables to pass into METplus. + + * If automated via a workflow manager, it is recommended the user consider + configuring the use cases to run smaller amounts of data. + +Where will METplus be run? Local machine, project machine, HPC system, in the cloud (e.g. AWS)? Serial runs or parallelized? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + + * Running on linux or a project machine – identify where METplus is + installed by running **which run_metplus.py**; it is recommended an + additional user.conf or system.conf file is passed into the + **run_metplus.py** to direct where output should be written. + + * Running on HPC systems - check with the system admin to see if it + has been configured as a module and how to load netCDF and Python + modules. For NOAA and NCAR HPCs systems, please refer to the + `Existing Builds `_ + pages for the desired version for instructions on how to load the METplus + related modules. + + * Running on Cloud (AWS) - these instructions are coming soon. + + * Running in parallel - As of MET v10.1.0 Grid-Stat can be run in parallel. + Please reach out via + `METplus Discussions `_ + if help is needed. + +Would a flowchart help to provide clarity? +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Utilizing a flowchart can assist in identifying which verification +steps can be completed by which METplus tools. + +.. _running-metplus: + +Running METplus +=============== + +Example Wrapper Use Case +------------------------ + +* Create a :ref:`user_configuration_file` + (named user_system.conf in this example). + +* Run the Example Wrapper use case. In a terminal, run:: + + run_metplus.py \ + /path/to/METplus/parm/use_cases/met_tool_wrapper/Example/Example.conf \ + /path/to/user_system.conf + +Replacing */path/to/user_system.conf* with the path to the +user configuration file and +*/path/to/METplus* with the path to the location where METplus is installed. + +The last line of the screen output should match this format:: + + 05/04 09:42:52.277 metplus (met_util.py:212) INFO: METplus has successfully finished running. + +If this log message is not shown, there is likely an issue with one or more +of the default configuration variable overrides in the +:ref:`user_configuration_file`. + +This use case does not utilize any of the MET tools, but simply demonstrates +how the :ref:`common_config_variables` control a use case run. + +If the run was successful, the line above the success message should contain +the path to the METplus log file that was generated:: + + 05/04 09:44:21.534 metplus (met_util.py:211) INFO: Check the log file for more information: /path/to/output/logs/metplus.log.20210504094421 + +* Review the log file and compare it to the Example.conf use case + configuration file to see how the settings correspond to the result. + +* Review the :ref:`metplus_final.conf` file to see all + of the settings that were used in the use case. + +GridStat Wrapper Basic Use Case +------------------------------- + +* :ref:`obtain_sample_input_data` for the **met_tool_wrapper** use cases. + The tarfile should be in the directory that corresponds to the + major/minor release and starts with sample_data-met_tool_wrapper. + +* Create a :ref:`user_configuration_file` (named user_system.conf in this + example). Ensure that **INPUT_BASE** is set + to the directory where the sample data tarfile was uncompressed. + +* Run the GridStat Wrapper basic use case. In a terminal, run:: + + run_metplus.py \ + /path/to/METplus/parm/use_cases/met_tool_wrapper/GridStat/GridStat.conf \ + /path/to/user_system.conf + +Replacing */path/to/user_system.conf* with the path to the +user configuration file and +*/path/to/METplus* with the path to the location where METplus is installed. + +If the run was successful, the line above the success message should contain +the path to the METplus log file that was generated. + +* Review the log file and compare it to the **GridStat.conf** use case + configuration file to see how the settings correspond to the result. + +* Review the :ref:`metplus_final.conf` file to see all + of the settings that were used in the use case. + diff --git a/docs/Users_Guide/index.rst b/docs/Users_Guide/index.rst index a1ab31578a..e08f2ea38d 100644 --- a/docs/Users_Guide/index.rst +++ b/docs/Users_Guide/index.rst @@ -78,6 +78,7 @@ is sponsored by NSF. :numbered: 4 overview + getting_started installation systemconfiguration wrappers diff --git a/docs/Users_Guide/installation.rst b/docs/Users_Guide/installation.rst index 3731817501..2a2d1ba090 100644 --- a/docs/Users_Guide/installation.rst +++ b/docs/Users_Guide/installation.rst @@ -1,8 +1,8 @@ .. _install: -********************* -Software Installation -********************* +************ +Installation +************ Introduction ============ diff --git a/docs/Users_Guide/systemconfiguration.rst b/docs/Users_Guide/systemconfiguration.rst index eb78d81b02..b823a8f9c2 100644 --- a/docs/Users_Guide/systemconfiguration.rst +++ b/docs/Users_Guide/systemconfiguration.rst @@ -1,8 +1,8 @@ .. _sysconf: -******************** -System Configuration -******************** +************* +Configuration +************* This chapter is a guide on configuring METplus Wrappers. @@ -31,13 +31,13 @@ Config Best Practices / Recommendations * If configuring METplus Wrappers in a common location for multiple users: - * It is recommended that the values for **MET_INSTALL_DIR** and - **INPUT_BASE** are changed to valid values in the - :ref:`default_configuration_file`. + * It is recommended that the values for **MET_INSTALL_DIR** and + **INPUT_BASE** are changed to valid values in the + :ref:`default_configuration_file`. - * It is recommended to leave **OUTPUT_BASE** set to the default value in - the :ref:`default_configuration_file`. This prevents multiple users from - accidentally writing to the same output directory. + * It is recommended to leave **OUTPUT_BASE** set to the default value in + the :ref:`default_configuration_file`. This prevents multiple users from + accidentally writing to the same output directory. * If obtaining the METplus Wrappers with the intention of updating the same local directory as new versions become available, @@ -534,79 +534,6 @@ given use case. More information about the variables set in the use case configuration files can be found in the :ref:`common_config_variables` section. -.. _running-metplus: - -Running METplus -=============== - -Example Wrapper Use Case ------------------------- - -* Create a :ref:`user_configuration_file` - (named user_system.conf in this example) - -* Run the Example Wrapper use case. In a terminal, run:: - - run_metplus.py \ - /path/to/METplus/parm/use_cases/met_tool_wrapper/Example/Example.conf \ - /path/to/user_system.conf - -replacing **/path/to/user_system.conf** with the path to the -user configuration file and -**/path/to/METplus** with the path to the location where METplus is installed - -The last line of the screen output should match this format:: - - 05/04 09:42:52.277 metplus (met_util.py:212) INFO: METplus has successfully finished running. - -If this log message is not shown, there is likely an issue with one or more -of the default configuration variable overrides in the -:ref:`user_configuration_file`. - -This use case does not utilize any of the MET tools, but simply demonstrates -how the :ref:`common_config_variables` control a use case run. - -If the run was successful, the line above the success message should contain -the path to the METplus log file that was generated:: - - 05/04 09:44:21.534 metplus (met_util.py:211) INFO: Check the log file for more information: /path/to/output/logs/metplus.log.20210504094421 - -* Review the log file and compare it to the Example.conf use case - configuration file to see how the settings correspond to the result. - -* Review the :ref:`metplus_final.conf` file to see all - of the settings that were used in the use case. - -GridStat Wrapper Basic Use Case -------------------------------- - -* :ref:`obtain_sample_input_data` for the **met_tool_wrapper** use cases. - The tarfile should be in the directory that corresponds to the - major/minor release and starts with sample_data-met_tool_wrapper - -* Create a :ref:`user_configuration_file` (named user_system.conf in this - example). Ensure that **INPUT_BASE** is set - to the directory where the sample data tarfile was uncompressed. - -* Run the GridStat Wrapper basic use case. In a terminal, run:: - - run_metplus.py \ - /path/to/METplus/parm/use_cases/met_tool_wrapper/GridStat/GridStat.conf \ - /path/to/user_system.conf - -replacing **/path/to/user_system.conf** with the path to the -user configuration file and -**/path/to/METplus** with the path to the location where METplus is installed - -If the run was successful, the line above the success message should contain -the path to the METplus log file that was generated. - -* Review the log file and compare it to the GridStat.conf use case - configuration file to see how the settings correspond to the result. - -* Review the :ref:`metplus_final.conf` file to see all - of the settings that were used in the use case. - .. _common_config_variables: Common Config Variables @@ -810,7 +737,7 @@ then three times will be processed: 2. Initialized on 2019-02-01 at 00Z / valid on 2019-02-01 at 06Z 3. Initialized on 2019-02-01 at 00Z / valid on 2019-02-01 at 09Z -You can also define :term:`LEAD_SEQ` using a special notation for many +The user can also define :term:`LEAD_SEQ` using a special notation for many forecast leads. The notation is **begin_end_incr(b,e,i)** where b = the first lead value, e = the last lead value (inclusive), and i = the increment between leads. For example:: @@ -828,9 +755,9 @@ the :term:`LEAD_SEQ` variable for the **SeriesByLead Wrapper Only**. If :term:`SERIES_BY_LEAD_GROUP_FCSTS` = True, then groups of forecast leads can be defined to be evaluated together. -You can define any number of these groups by setting +Any number of these groups can be defined by setting configuration variables LEAD_SEQ_1, LEAD_SEQ_2, ..., :term:`LEAD_SEQ_\`. -You can define the value with a +The value can be defined with a comma-separated list of integers (currently only hours are supported here) or using the special begin_end_incr(b,e,i) notation described just above. Each :term:`LEAD_SEQ_\` must have a corresponding @@ -941,7 +868,7 @@ Example 3:: This will skip every 30th and 31st day **and** every 3rd month. -You can use **begin_end_incr(b,e,i)** syntax to define a range of times to +**begin_end_incr(b,e,i)** syntax can be used to define a range of times to skip. b = begin value, e = end value, @@ -959,7 +886,7 @@ This is equivalent to:: [config] SKIP_TIMES = "%H:0,2,4,6,8,10,12,14,16,18,20,22" -You can also specify multiple strftime directives in a single time format. +Multiple strftime directives can be specified in a single time format. Example 5:: @@ -997,15 +924,15 @@ and the METplus Wrappers is run with:: [config] VALID_END = {now?fmt=%Y%m%d%H} -then the value of :term:`VALID_END` will be set to 2019042608. You can also -use {today} to substitute the -current YYYYMMDD, i.e. 20190426. You cannot change the formatting for -the 'today' keyword. +then the value of :term:`VALID_END` will be set to 2019042608. {today} +can also be used to substitute the +current YYYYMMDD, i.e. 20190426. The formatting for +the 'today' keyword cannot be changed.. Shift Keyword """"""""""""" -You can use the 'shift' keyword to shift the current time by any number of +The 'shift' keyword can be used to shift the current time by any number of seconds. For example, if the METplus Wrappers are run at the same clock time with:: @@ -1029,7 +956,7 @@ METplus Wrappers will process the following valid times: Truncate Keyword """""""""""""""" -You may want to configure the METplus Wrappers to process at 00Z, 06Z, 12Z, +The user may want to configure the METplus Wrappers to process at 00Z, 06Z, 12Z, and 18Z of a given day instead of 02Z, 08Z, 14Z, and 20Z. Having to adjust the shift amount differently if running at 08Z or 09Z to get the times to line up would be tedious. Instead, use the 'truncate' keyword. @@ -1705,8 +1632,8 @@ Shifting Times in Filename Templates Users can use the 'shift' keyword to adjust the time referenced in the filename template relative to the run time. For example, if the input files used contained data from 01Z on the date specified in the filename to 01Z on -the following day. In this example, for a run at 00Z you want to use the -file from the previous day and for the 01Z to 23Z runs you want to use the +the following day. In this example, for a run at 00Z use the +file from the previous day and for the 01Z to 23Z runs, use the file that corresponds to the current day. Here is an example:: [config] @@ -1724,7 +1651,7 @@ Using Windows to find Valid Files The [FCST/OBS]_FILE_WINDOW_[BEGIN/END] configuration variables can be used if the time information in the input data does not exactly line up with the -run time but you still want to process the data. The default value of the +run time but the user still wants to process the data. The default value of the file window begin and end variables are both 0 seconds. If both values are set to 0, METplus Wrappers will require that a file matching the template with the exact time requested exists. If either value is non-zero, METplus @@ -1777,7 +1704,7 @@ Wrapper Specific Windows ^^^^^^^^^^^^^^^^^^^^^^^^ A user may need to specify a different window on a wrapper-by-wrapper basis. -If this is the case, you can override the file window values for each +If this is the case, the user can override the file window values for each wrapper. Consider the following configuration:: [config] @@ -1801,7 +1728,7 @@ Runtime Frequency Some wrappers have an option to specify how frequently to process data. It can be run once to process all of the available files in the desired time range, -or it can be configured to run over different intervals. This allows you to +or it can be configured to run over different intervals. This allows the user to aggregate the output in a variety of ways. The wrappers that support this functionality (along with the configuration variable that controls the setting) include: @@ -2006,12 +1933,12 @@ supports METplus configuration variables that control the MET tool configuration file settings. **The METplus wrappers provide a special "wrapped" MET configuration file that references environment variables that are set by the wrappers based on the -values set in the METplus configuration files. YOU SHOULD NOT SET ANY OF THESE -ENVIRONMENT VARIABLES YOURSELF! THEY WILL BE OVERWRITTEN BY METPLUS WHEN IT +values set in the METplus configuration files. THE USER SHOULD NOT SET ANY OF +THESE ENVIRONMENT VARIABLES! THEY WILL BE OVERWRITTEN BY METPLUS WHEN IT CALLS THE MET TOOLS!** If there is a setting in the MET configuration file that is not currently -supported by METplus you'd like to control, please refer to: +supported by METplus that the user would like to control, please refer to: :ref:`Overriding Unsupported MET config file settings`. The following section demonstrates a few examples using GridStat. @@ -2661,7 +2588,7 @@ This is the equivalent of running this bash command:: on the command line before calling run_metplus.py. -You can also reference other variables in the METplus config file. +Other variables can also be referenced in the METplus config file. For example:: [config] @@ -2686,7 +2613,7 @@ on the command line before calling run_metplus.py. Setting Config Variables with Environment Variables =================================================== -You can set METplus config variables to the value of local environment +The METplus config variables can be set to the value of local environment variables when METplus is run. To set any METplus config variable to the value of a local environment variable, use the following syntax:: @@ -2747,10 +2674,10 @@ Starting in METplus 3.0, users are required to either explicitly set both FCST_* Example:: - (met_util.py) ERROR: If FCST_VAR1_NAME is set, you must either set OBS_VAR1_NAME or change FCST_VAR1_NAME to BOTH_VAR1_NAME - (met_util.py) ERROR: If FCST_VAR2_NAME is set, you must either set OBS_VAR2_NAME or change FCST_VAR2_NAME to BOTH_VAR2_NAME - (met_util.py) ERROR: If FCST_VAR1_LEVELS is set, you must either set OBS_VAR1_LEVELS or change FCST_VAR1_LEVELS to BOTH_VAR1_LEVELS - (met_util.py) ERROR: If FCST_VAR2_LEVELS is set, you must either set OBS_VAR2_LEVELS or change FCST_VAR2_LEVELS to BOTH_VAR2_LEVELS + (met_util.py) ERROR: If FCST_VAR1_NAME is set, the user must either set OBS_VAR1_NAME or change FCST_VAR1_NAME to BOTH_VAR1_NAME + (met_util.py) ERROR: If FCST_VAR2_NAME is set, the user must either set OBS_VAR2_NAME or change FCST_VAR2_NAME to BOTH_VAR2_NAME + (met_util.py) ERROR: If FCST_VAR1_LEVELS is set, the user must either set OBS_VAR1_LEVELS or change FCST_VAR1_LEVELS to BOTH_VAR1_LEVELS + (met_util.py) ERROR: If FCST_VAR2_LEVELS is set, the user must either set OBS_VAR2_LEVELS or change FCST_VAR2_LEVELS to BOTH_VAR2_LEVELS These cases can be handled automatically by using the :ref:`validate_config`, but users should review the suggested changes, as they may want to update differently. @@ -2809,7 +2736,7 @@ Example log output:: (met_util.py) ERROR: Set GRID_STAT_CLIMO_MEAN_INPUT_[DIR/TEMPLATE] in a METplus config file to set CLIMO_MEAN_FILE in a MET config (met_util.py) ERROR: output_prefix variable should reference ${OUTPUT_PREFIX} environment variable - (met_util.py) INFO: You will need to add GRID_STAT_OUTPUT_PREFIX to the METplus config file that sets GRID_STAT_CONFIG_FILE. Set it to: + (met_util.py) INFO: GRID_STAT_OUTPUT_PREFIX will need to be added to the METplus config file that sets GRID_STAT_CONFIG_FILE. Set it to: (met_util.py) INFO: GRID_STAT_OUTPUT_PREFIX = {CURRENT_FCST_NAME}_vs_{CURRENT_OBS_NAME} These cases can be handled automatically by using the :ref:`validate_config`, but users should review the suggested changes and make sure they add the appropriate recommended METplus configuration variables to their files to achieve the same behavior. @@ -2843,7 +2770,7 @@ The script named validate_config.py is found in the same directory as run_metplu run_metplus.py ./my_conf.py ./another_config.py validate_config.py ./my_conf.py ./another_config.py -You must pass a valid configuration to the script, as in you must properly set :term:`MET_INSTALL_DIR`, :term:`INPUT_BASE`, and :term:`OUTPUT_BASE`, or it will not run. +A valid configuration must be passed to the script, as in the user must properly set :term:`MET_INSTALL_DIR`, :term:`INPUT_BASE`, and :term:`OUTPUT_BASE`, or it will not run. The script will evaluate all of the configuration files, including any MET configuration file that is referenced in a _CONFIG_FILE variable, such as :term:`GRID_STAT_CONFIG_FILE`. For each deprecated item that is found, the script will suggest a replacement for the file where the deprecated item was found. @@ -2912,7 +2839,7 @@ Example 5 (Another MET Configuration File):: Would you like the make this change to DeprecatedConfig? (y/n)[n] - IMPORTANT: If it is not already set, add the following in the [config] section to your METplus configuration file that sets GRID_STAT_CONFIG_FILE: + IMPORTANT: If it is not already set, add the following in the [config] section to the METplus configuration file that sets GRID_STAT_CONFIG_FILE: GRID_STAT_OUTPUT_PREFIX = {CURRENT_FCST_NAME}_vs_{CURRENT_OBS_NAME} Make this change before continuing! [OK] @@ -2921,7 +2848,7 @@ Example 5 (Another MET Configuration File):: While the METplus developers are very diligent to include deprecated variables in this functionality, some may slip through the cracks. When upgrading to a new version of METplus, it is important to test - and review your use cases to ensure they produce the same results as + and review the use cases to ensure they produce the same results as the previous version. Please create a post in the `METplus GitHub Discussions Forum `_ with any questions.