Merge pull request #59 from kthyng/more_general

Making the codebase more general

.pre-commit-config.yaml

@@ -51,7 +51,7 @@ repos:
   rev: v0.982
   hooks:
   - id: mypy
-    additional_dependencies: [types-setuptools, types-PyYAML, types-requests]
+    additional_dependencies: [types-setuptools, types-PyYAML, types-requests, types-DateTimeRange]
     exclude: docs/source/conf.py
     args: [--ignore-missing-imports]
 

ci/environment-py3.10.yml

@@ -6,13 +6,16 @@ dependencies:
   ############## These will have to be adjusted to your specific project
   # - cf_pandas
   - cf_xarray
+  - cmocean
+  - datetimerange
   - extract_model
   # - intake
   # - intake-axds
   # - intake-erddap
   - matplotlib-base
   - numpy
   - pandas
+  - pyproj
   - scipy
   - xarray
   ##############

ci/environment-py3.8.yml

@@ -6,13 +6,16 @@ dependencies:
   ############## These will have to be adjusted to your specific project
   # - cf_pandas
   - cf_xarray
+  - cmocean
+  - datetimerange
   - extract_model
   # - intake
   # - intake-axds
   # - intake-erddap
   - matplotlib-base
   - numpy
   - pandas
+  - pyproj
   - scipy
   - xarray
   ##############

ci/environment-py3.9.yml

@@ -6,13 +6,16 @@ dependencies:
   ############## These will have to be adjusted to your specific project
   # - cf_pandas
   - cf_xarray
+  - cmocean
+  - datetimerange
   - extract_model
   # - intake
   # - intake-axds
   # - intake-erddap
   - matplotlib-base
   - numpy
   - pandas
+  - pyproj
   - scipy
   - xarray
   ##############

docs/environment.yml

@@ -9,6 +9,8 @@ dependencies:
    - cartopy
   #  - cf_pandas
    - cf_xarray
+   - cmocean
+   - datetimerange
    - extract_model
    - intake
   #  - intake-axds
@@ -20,6 +22,7 @@ dependencies:
    # These are needed for the docs themselves
    - jupytext
    - numpydoc
+   - pyproj
    - requests
    - sphinx
    - sphinx-markdown-tables

docs/index.rst

@@ -43,6 +43,7 @@ To install from PyPI:
    :caption: For contributors
 
    developer.md
+   whats_new
    GitHub repository <https://github.com/axiom-data-science/ocean-model-skill-assessor>
 
 

docs/whats_new.md

@@ -0,0 +1,5 @@
+# What's New
+
+## unreleased version
+
+* `omsa.run` now saves the polygon found for the input model into the project directory.

environment.yml

@@ -8,14 +8,18 @@ dependencies:
   - aiohttp
   # - cf_pandas
   - cf_xarray
+  - cmocean
+  - datetimerange
   - extract_model
   - intake
   # - intake-axds
   # - intake-erddap
   - intake-xarray
   - matplotlib-base
+  - oceans
   - pandas
   - pip
+  - pyproj
   - requests
   - scipy
   - xarray

ocean_model_skill_assessor/accessor.py

@@ -2,42 +2,168 @@
 Class to facilitate some functions directly on DataFrames.
 """
 
-from pandas import DatetimeIndex
+import pandas as pd
+import xarray as xr
+
+# from pandas import DatetimeIndex
 from pandas.api.extensions import register_dataframe_accessor
 
-from ocean_model_skill_assessor.plot import time_series
+from ocean_model_skill_assessor.plot import line, scatter, surface
 
 from .stats import compute_stats
 
 
 @register_dataframe_accessor("omsa")
+@xr.register_dataset_accessor("omsa")
 class SkillAssessorAccessor:
     """Class to facilitate some functions directly on DataFrames."""
 
-    def __init__(self, df):
+    def __init__(self, dd):
         """
+
+        don't validate with this because might not be time series
+
         Parameters
         ----------
-        pandas_obj: DataFrame
-            Should be observations, could be unloaded dask dataframe
+        obj: DataFrame or Dataset.
+            Should be observations and model, could be unloaded dask dataframe
         """
-        self._validate(df)
-        self.df = df
+        # if isinstance(dd, pd.DataFrame):
+        #     self._validate(dd)
+        self.dd = dd
 
     @staticmethod
-    def _validate(df):
+    def _validate(dd):
         """DataFrame must have datetimes as index."""
-        if not isinstance(df.index, DatetimeIndex):
+        if not isinstance(dd.index, pd.DatetimeIndex):
             raise TypeError("DataFrame index must be datetimes")
 
     @property
     def compute_stats(self):
         """Run `compute_stats` on DataFrame."""
         if not hasattr(self, "_compute_stats"):
-            stats = compute_stats(self.df["obs"], self.df["model"])
+            stats = compute_stats(self.dd["obs"], self.dd["model"])
             self._compute_stats = stats
         return self._compute_stats
 
-    def plot(self, **kwargs):
+    def plot(self, featuretype=None, key_variable=None, **kwargs):
         """Plot."""
-        time_series.plot(self.df["obs"], self.df["model"], **kwargs)
+        # import pdb; pdb.set_trace()
+        import xcmocean
+
+        # cmap and cmapdiff
+        da = xr.DataArray(name=key_variable)
+
+        # use featuretype to determine plot type, otherwise assume time series
+        with xr.set_options(cmap_sequential=da.cmo.seq, cmap_divergent=da.cmo.div):
+            if featuretype is not None:
+                if featuretype == "timeSeries":
+                    xname, yname, zname = (
+                        self.dd.index.name or "index",
+                        ["obs", "model"],
+                        None,
+                    )
+                    xlabel, ylabel = "", key_variable
+                    # xname, yname, zname = self.dd.cf["T"].name, ["obs","model"], None
+                    # import pdb; pdb.set_trace()
+                    line.plot(
+                        self.dd.reset_index(),
+                        xname,
+                        yname,
+                        xlabel=xlabel,
+                        ylabel=ylabel,
+                        figsize=(15, 5),
+                        **kwargs
+                    )
+                elif featuretype == "trajectoryProfile":
+                    xname, yname, zname = "distance", "Z", ["obs", "model"]
+                    surface.plot(
+                        self.dd.reset_index(),
+                        xname,
+                        yname,
+                        zname,
+                        kind="scatter",
+                        **kwargs
+                    )
+                    # surface.plot(xname, yname, self.dd["obs"], self.dd["model"], kind="scatter", **kwargs)
+                    # scatter.plot(self.dd["obs"], self.dd["model"], **kwargs)
+                elif featuretype == "timeSeriesProfile":
+                    xname, yname, zname = "T", "Z", ["obs", "model"]
+                    surface.plot(
+                        self.dd.reset_index(),
+                        xname,
+                        yname,
+                        zname,
+                        kind="pcolormesh",
+                        **kwargs
+                    )
+                    # surface.plot(xname, yname, self.dd["obs"].squeeze(), self.dd["model"].squeeze(), **kwargs)
+                elif featuretype == "profile":
+                    # use transpose so that index depth is plotted on y axis instead of x axis
+                    # import pdb; pdb.set_trace()
+                    xname, yname, zname = (
+                        ["obs", "model"],
+                        self.dd.index.name or "index",
+                        None,
+                    )
+                    # import pdb; pdb.set_trace()
+                    xlabel, ylabel = key_variable, yname
+                    # xname, yname, zname = ["obs","model"], self.dd.cf["Z"].name, None
+                    line.plot(
+                        self.dd.reset_index(),
+                        xname,
+                        yname,
+                        xlabel=xlabel,
+                        ylabel=ylabel,
+                        figsize=(4, 8),
+                        **kwargs
+                    )
+            else:
+                xname, yname, zname = "index", ["obs", "model"], None
+                line.plot(
+                    self.dd.reset_index(), xname, yname, figsize=(15, 5), **kwargs
+                )
+                # time_series.plot(self.dd["obs"], self.dd["model"], **kwargs)
+
+
+# @xr.register_dataset_accessor("omsa")
+# class SkillAssessorAccessor:
+#     """Class to facilitate some functions directly on Datasets."""
+
+#     def __init__(self, ds):
+#         """
+#         Parameters
+#         ----------
+#         xarray_obj: Dataset
+#             Should be observations and model, could be unloaded dask dataframe
+#         """
+#         self._validate(ds)
+#         self.df = ds
+
+#     # @staticmethod
+#     # def _validate(df):
+#     #     """DataFrame must have datetimes as index."""
+#     #     if not isinstance(df.index, DatetimeIndex):
+#     #         raise TypeError("DataFrame index must be datetimes")
+
+#     @property
+#     def compute_stats(self):
+#         """Run `compute_stats` on Dataset"""
+#         if not hasattr(self, "_compute_stats"):
+#             stats = compute_stats(self.ds["obs"], self.ds["model"])
+#             self._compute_stats = stats
+#         return self._compute_stats
+
+#     def plot(self, featuretype=None, **kwargs):
+#         """Plot."""
+
+#         # use featuretype to determine plot type, otherwise assume time series
+#         if featuretype is not None:
+#             if featuretype == "timeSeries":
+#                 time_series.plot(self.df["obs"], self.df["model"], **kwargs)
+#             elif featuretype == "trajectoryProfile":
+#                 scatter.plot(self.df["obs"], self.df["model"], **kwargs)
+#             elif featuretype == "timeSeriesProfile":
+#                 surface.plot(self.df["obs"], self.df["model"], **kwargs)
+#         else:
+#             time_series.plot(self.df["obs"], self.df["model"], **kwargs)