Merge pull request #640 from MetOffice/639_surface_field_histogram_ex…

…tension_to_453

Surface field histogram

cset-workflow/includes/deterministic_domain_histogram_series.cylc

@@ -0,0 +1,16 @@
+{% if DOMAIN_HISTOGRAM_SERIES_FIELD %}
+[runtime]
+{% for model_field in MODEL_LEVEL_MODEL_FIELDS %}
+    [[pre_process_deterministic_domain_histogram_series_{{model_field}}]]
+    inherit = PARALLEL
+        [[[environment]]]
+        CSET_RECIPE_NAME = "generic_histogram_series.yaml"
+        CSET_ADDOPTS = "--VARNAME='{{model_field}}' --MLEVEL='{{UM_MODEL_LEVELS}}'"
+
+    [[collate_deterministic_domain_histogram_series_{{model_field}}]]
+    inherit = COLLATE
+        [[[environment]]]
+        CSET_RECIPE_NAME = "generic_histogram_series.yaml"
+        CSET_ADDOPTS = "--VARNAME='{{model_field}}' --MLEVEL='{{UM_MODEL_LEVELS}}'"
+{% endfor %}
+{% endif %}

cset-workflow/includes/deterministic_domain_surface_histogram_series.cylc

@@ -0,0 +1,16 @@
+{% if DOMAIN_SURFACE_HISTOGRAM_SERIES_FIELD %}
+[runtime]
+{% for model_field in SURFACE_MODEL_FIELDS %}
+    [[pre_process_deterministic_domain_surface_histogram_series_{{model_field}}]]
+    inherit = PARALLEL
+        [[[environment]]]
+        CSET_RECIPE_NAME = "generic_surface_histogram_series.yaml"
+        CSET_ADDOPTS = "--VARNAME='{{model_field}}'"
+
+    [[collate_deterministic_domain_surface_histogram_series_{{model_field}}]]
+    inherit = COLLATE
+        [[[environment]]]
+        CSET_RECIPE_NAME = "generic_surface_histogram_series.yaml"
+        CSET_ADDOPTS = "--VARNAME='{{model_field}}'"
+{% endfor %}
+{% endif %}

cset-workflow/meta/rose-meta.conf

@@ -470,6 +470,14 @@ type=python_boolean
 compulsory=true
 sort-key=surface2
 
+[template variables=DOMAIN_SURFACE_HISTOGRAM_SERIES_FIELD]
+ns=Diagnostics
+description=Create a series of histogram plots for selected surface fields for each cycle time.
+help=See includes/deterministic_domain_surface_histogram_series.cylc. This diagnostic requires the user to also enable and enter the SURFACE_MODEL_FIELDS. The series_coordinate in the recipe is set to "time", but can be switched to any other coordinate.
+type=python_boolean
+compulsory=true
+sort-key=surface2
+
 [template variables=LFRIC_PLOT_SPATIAL_SURFACE_MODEL_FIELD]
 ns=Diagnostics
 description=Create plots for the specified surface fields for structured LFRic data.
@@ -602,6 +610,23 @@ type=python_list
 compulsory=true
 sort-key=model1
 
+[template variables=LFRIC_PRESSURE_LEVEL_MODEL_FIELDS]
+ns=Diagnostics
+description=List of standard names of model fields on pressure levels to plot.
+help=Include a list of variable names in python list format, e.g: ["var1","var2"]
+type=python_list
+compulsory=true
+sort-key=pressure3
+
+[template variables=LFRIC_PRESSURE_LEVELS]
+ns=Diagnostics
+description=List of pressure levels to generate plots for.
+help=Include an list of integer pressure levels in hPa in python list format, e.g: [1000,850]
+type=python_list
+compulsory=true
+sort-key=pressure3
+
+
 [template variables=PLOT_SPATIAL_PRESSURE_LEVEL_MODEL_FIELD]
 ns=Diagnostics
 description=Create plots for the specified pressure level fields.

src/CSET/operators/plot.py

@@ -411,8 +411,8 @@ def _plot_and_save_vertical_line_series(
         # Set y-axis limits and ticks.
         ax.set_ylim(1100, 100)
 
-    # test if series_coordinate is model level data. The um data uses model_level_number
-    # and lfric uses full_levels as coordinate.
+    # Test if series_coordinate is model level data. The UM data uses
+    # model_level_number and lfric uses full_levels as coordinate.
     elif series_coordinate in ("model_level_number", "full_levels", "half_levels"):
         # Define y-ticks and labels for vertical axis.
         y_ticks = cube.coord(series_coordinate).points
@@ -431,6 +431,9 @@ def _plot_and_save_vertical_line_series(
         xlabel=f"{cube.name()} / {cube.units}",
         title=title,
     )
+    ax.ticklabel_format(axis="x")
+    ax.tick_params(axis="y")
+    ax.autoscale()
 
     # Save plot.
     fig.savefig(filename, bbox_inches="tight", dpi=150)
@@ -725,6 +728,7 @@ def spatial_contour_plot(
     except iris.exceptions.CoordinateNotFoundError:
         pass
 
+    # Must have a sequence coordinate.
     try:
         cube.coord(sequence_coordinate)
     except iris.exceptions.CoordinateNotFoundError as err:
@@ -876,23 +880,24 @@ def plot_vertical_line_series(
     except iris.exceptions.CoordinateNotFoundError as err:
         raise ValueError(f"Cube must have a {series_coordinate} coordinate.") from err
 
-    # If several individual vertical lines are plotted with time as sequence_coordinate
-    # for the time slider option.
     try:
-        cube.coord(sequence_coordinate)
+        if cube.ndim > 1:
+            cube.coord(sequence_coordinate)
     except iris.exceptions.CoordinateNotFoundError as err:
-        raise ValueError(f"Cube must have a {sequence_coordinate} coordinate.") from err
+        raise ValueError(
+            f"Cube must have a {sequence_coordinate} coordinate or be 1D."
+        ) from err
 
     # Ensure we have a name for the plot file.
     recipe_title = get_recipe_metadata().get("title", "Untitled")
     if filename is None:
         filename = slugify(recipe_title)
 
-    # set the lower and upper limit for the x-axis to ensure all plots
-    # have same range. This needs to read the whole cube over the range of
-    # the sequence and if applicable postage stamp coordinate.
-    # This only works if the plotting is done in the collate section of a
-    # recipe and not in the parallel section of a recipe.
+    # Set the lower and upper limit for the x-axis to ensure all plots have same
+    # range. This needs to read the whole cube over the range of the sequence
+    # and if applicable postage stamp coordinate. This only works if the
+    # plotting is done in the collate section of a recipe and not in the
+    # parallel section of a recipe.
     vmin = np.floor(cube.data.min())
     vmax = np.ceil(cube.data.max())
 
@@ -1032,10 +1037,10 @@ def plot_histogram_series(
 
     A histogram plot can be plotted, but if the sequence_coordinate (i.e. time)
     is present then a sequence of plots will be produced using the time slider
-    functionality to scroll through histograms against time.
-    If a stamp_coordinate is present then postage stamp plots will be produced.
-    If stamp_coordinate and single_plot is True,
-    all postage stamp plots will be plotted in a single plot instead of separate postage stamp plots.
+    functionality to scroll through histograms against time. If a
+    stamp_coordinate is present then postage stamp plots will be produced. If
+    stamp_coordinate and single_plot is True, all postage stamp plots will be
+    plotted in a single plot instead of separate postage stamp plots.
 
     Parameters
     ----------
@@ -1047,18 +1052,19 @@ def plot_histogram_series(
         to the recipe name.
     sequence_coordinate: str, optional
         Coordinate about which to make a plot sequence. Defaults to ``"time"``.
-        This coordinate must exist in the cube and will be used for the time slider.
+        This coordinate must exist in the cube and will be used for the time
+        slider.
     stamp_coordinate: str, optional
         Coordinate about which to plot postage stamp plots. Defaults to
         ``"realization"``.
     single_plot: bool, optional
-        If True, all postage stamp plots will be plotted in a single plot.
-        If False, each postage stamp plot will be plotted separately.
-        Is only valid if stamp_coordinate exists and has more than a single point.
+        If True, all postage stamp plots will be plotted in a single plot. If
+        False, each postage stamp plot will be plotted separately. Is only valid
+        if stamp_coordinate exists and has more than a single point.
     histtype: str, optional
-        The type of histogram to plot. Options are "step" for a line
-        histogram or "barstacked", "stepfilled". "Step" is the default option,
-        but can be changed in the rose-suite.conf configuration.
+        The type of histogram to plot. Options are "step" for a line histogram
+        or "barstacked", "stepfilled". "Step" is the default option, but can be
+        changed in the rose-suite.conf configuration.
 
     Returns
     -------
@@ -1105,11 +1111,11 @@ def plot_histogram_series(
     except iris.exceptions.CoordinateNotFoundError as err:
         raise ValueError(f"Cube must have a {sequence_coordinate} coordinate.") from err
 
-    # Set the lower and upper limit for the colorbar to ensure all plots
-    # have same range. This needs to read the whole cube over the range of
-    # the sequence and if applicable postage stamp coordinate.
-    # This only works if the plotting is done in the collate section of a
-    # recipe and not in the parallel section of a recipe.
+    # Set the lower and upper limit for the colorbar to ensure all plots have
+    # same range. This needs to read the whole cube over the range of the
+    # sequence and if applicable postage stamp coordinate. This only works if
+    # the plotting is done in the collate section of a recipe and not in the
+    # parallel section of a recipe.
     vmin = np.floor((cube.data.min()))
     vmax = np.ceil((cube.data.max()))
 

src/CSET/recipes/generic_histogram_series.yaml

@@ -1,7 +1,13 @@
 category: Quick Look
 title: $VARNAME $MLEVEL Level Histogram Plot
 description: |
-  Extracts and plots the probability density of $MLEVEL level $VARNAME from a file at model level $MLEVEL. It uses [`plt.hist`](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.hist.html) to plot the probability density so that the area under the histogram integrates to 1. `stacked` is set to True so the sum of the histograms is normalized to 1. In case of ensemble data choose from postage stamp plot or single plot via the single_plot option in the recipe directly.
+  Extracts and plots the probability density of $MLEVEL level $VARNAME from a
+  file at model level $MLEVEL. It uses
+  [`plt.hist`](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.hist.html)
+  to plot the probability density so that the area under the histogram
+  integrates to 1. `stacked` is set to True so the sum of the histograms is
+  normalized to 1. In case of ensemble data choose from postage stamp plot or
+  single plot via the single_plot option in the recipe directly.
 
 parallel:
   - operator: read.read_cube
@@ -34,4 +40,3 @@ collate:
     # stamp_coordinate and single_plot optional and only required for ensemble data
     stamp_coordinate: "realization"
     single_plot: False
-    histtype: $HISTTYPE

src/CSET/recipes/generic_surface_histogram_series.yaml

@@ -0,0 +1,37 @@
+category: Quick Look
+title: $VARNAME Surface Level Histogram Plot
+description: |
+  Extracts and plots the probability density of surface `$VARNAME`. It uses
+  [`plt.hist`](https://matplotlib.org/stable/api/_as_gen/matplotlib.pyplot.hist.html)
+  to plot the probability density so that the area under the histogram
+  integrates to 1. stacked is set to True so the sum of the histograms is
+  normalized to 1.
+
+parallel:
+  - operator: read.read_cube
+    constraint:
+      operator: constraints.combine_constraints
+      variable_constraint:
+        operator: constraints.generate_var_constraint
+        varname: $VARNAME
+      validity_time_constraint:
+        operator: constraints.generate_time_constraint
+        time_start: $VALIDITY_TIME
+      pressure_level_constraint:
+        operator: constraints.generate_level_constraint
+        coordinate: pressure
+        levels: []
+
+  - operator: write.write_cube_to_nc
+    filename: intermediate/histogram
+
+
+collate:
+  - operator: read.read_cube
+    filename_pattern: intermediate/*.nc
+
+  - operator: write.write_cube_to_nc
+    overwrite: True
+
+  - operator: plot.plot_histogram_series
+    sequence_coordinate: time

src/CSET/recipes/lfric_generic_domain_mean_vertical_profile_series.yaml

@@ -1,6 +1,6 @@
 category: Quick Look
 title: Domain mean $VARNAME vertical profile as series
-description: Plots a time series of the vertical profile of domain mean $VARNAME using the model_level_number coordinate.
+description: Plots a time series of the vertical profile of domain mean $VARNAME using the full_levels coordinate.
 
 # Pre-processing steps.
 parallel:
@@ -27,8 +27,8 @@ parallel:
     filename: intermediate/model_level_domain_mean
 
 # Collation steps.
-# Reads in intermediate cube and plots it.
 collate:
+  # Reads in intermediate cube and plots it.
   - operator: read.read_cube
     filename_pattern: intermediate/*.nc
 
@@ -37,6 +37,6 @@ collate:
     series_coordinate: full_levels
     sequence_coordinate: time
 
-# Make a single NetCDF with all the data inside it.
+  # Make a single NetCDF with all the data inside it.
   - operator: write.write_cube_to_nc
     overwrite: True