discontinuum 1.0.1__py3-none-any.whl → 1.0.3__py3-none-any.whl

discontinuum/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '1.0.1'
-__version_tuple__ = version_tuple = (1, 0, 1)
+__version__ = version = '1.0.3'
+__version_tuple__ = version_tuple = (1, 0, 3)

{discontinuum-1.0.1.dist-info → discontinuum-1.0.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: discontinuum
-Version: 1.0.1
+Version: 1.0.3
 Summary: Estimate discontinuous timeseries from continuous covariates.
 Maintainer-email: Timothy Hodson <thodson@usgs.gov>
 License: License

{discontinuum-1.0.1.dist-info → discontinuum-1.0.3.dist-info}/RECORD

@@ -1,5 +1,5 @@
 discontinuum/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-discontinuum/_version.py,sha256=2Cl9kcQGg7NiH_eHRcv8wDYSW0X4uS5J0mGdZFO9-h0,511
+discontinuum/_version.py,sha256=jWOT7x0QtPxvOokiMtAg9lJ_WkkL0NXT4lwG0yqXne0,511
 discontinuum/data_manager.py,sha256=LiZoPR0nnu7YAUfh5L1ZDRfaS3dgfVIELXIHkzUKyBg,4416
 discontinuum/pipeline.py,sha256=1avuZnFai-b3HmihcpZ8M3WFNQ8lXAFSNTrnfl2NrY0,10074
 discontinuum/plot.py,sha256=eZQS6-Ydq8FFcEukPtNuDVB-weV6lHyWMyJ1hqTkVrU,2969
@@ -11,7 +11,7 @@ discontinuum/engines/pymc.py,sha256=phbtE-3UCSVcP1MhbXwAHIWDZWDr56wK9U7aRt-w-2o,
 discontinuum/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 discontinuum/providers/base.py,sha256=Yn2EHS1b4fYl09-m2MYuf2P9VRUXAP-WDpSoZrCbRvY,720
 discontinuum/tests/test_pipeline.py,sha256=_FhkGxbFIxNb35lGaIdZk7Zjgs6CkxEF3gFUX3PE8EU,918
-discontinuum-1.0.1.dist-info/licenses/LICENSE.md,sha256=XElVHHnS2uQ15M_Z2giPH1vmeWMzdpGQ48ItkuZurVA,1650
+discontinuum-1.0.3.dist-info/licenses/LICENSE.md,sha256=XElVHHnS2uQ15M_Z2giPH1vmeWMzdpGQ48ItkuZurVA,1650
 loadest_gp/__init__.py,sha256=YISfvbc7Zy2y0BOxS1A2KzqxyoNJTz0EnLMnRW6iVT8,740
 loadest_gp/plot.py,sha256=x2PK7vBCc44dX9lu5YV-rvw1u4pvXSLdcrTSvYLiHMA,2595
 loadest_gp/utils.py,sha256=m5QaqR_0JiuRXPfryH8nI5lODp8PqvQla5C05WDN3LY,2772
@@ -25,11 +25,11 @@ rating_gp/pipeline.py,sha256=1HgxN6DD3ZL5lhUb3DK2in2IXiml7W4Ja272GBMTc08,1884
 rating_gp/plot.py,sha256=CJphwqWWAfIY22j5Oz5DRwj7TcQCRyIQvM79_3KEdlc,9635
 rating_gp/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 rating_gp/models/base.py,sha256=e2Kq644I88YLHWPNA0qyRgitF5wimdLW4618vKX-o_s,1474
-rating_gp/models/gpytorch.py,sha256=rSxuTMoSeGK2LwqANMDMrmxDsOCXyAWSmENm6KFjRZ0,5930
+rating_gp/models/gpytorch.py,sha256=OUMKIdBgPjAxInYttKzGN2ou2zlc9V7BTUf4SuTVvwY,6043
 rating_gp/models/kernels.py,sha256=3xg2mhY3aEgjI3r5vyAll9MA4c3M5UKqRi3FApNhJJQ,11579
 rating_gp/providers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-rating_gp/providers/usgs.py,sha256=kgsgqHB90d3Ql8YvOeeXL_kmJdOzJ1uo-f29-PupWg8,6909
-discontinuum-1.0.1.dist-info/METADATA,sha256=7jkJhnAoeiUP8XIGuLtF5Ws79QoQ1Jc2pCNtnc6DSK8,6231
-discontinuum-1.0.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-discontinuum-1.0.1.dist-info/top_level.txt,sha256=mwU_PSFrZYSJrBgqIuTJTo7Pp9ODDv6XdDed7kAagXM,34
-discontinuum-1.0.1.dist-info/RECORD,,
+rating_gp/providers/usgs.py,sha256=KmKYN3c8Mi-ly2l6X80WT3taEhqCPXeEcRNi9HvbJmY,8134
+discontinuum-1.0.3.dist-info/METADATA,sha256=RukIH_49qEGFmjgaqaN2wTUwR4SPpD9zUWeHNhPosZI,6231
+discontinuum-1.0.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+discontinuum-1.0.3.dist-info/top_level.txt,sha256=mwU_PSFrZYSJrBgqIuTJTo7Pp9ODDv6XdDed7kAagXM,34
+discontinuum-1.0.3.dist-info/RECORD,,

rating_gp/models/gpytorch.py

@@ -71,7 +71,8 @@ class RatingGPMarginalGPyTorch(
         # noise, *and* you did not specify noise. This is treated as a no-op."
         self.likelihood = gpytorch.likelihoods.FixedNoiseGaussianLikelihood(
             noise=noise,
-            learn_additional_noise=False,
+            learn_additional_noise=True,
+            noise_prior=gpytorch.priors.HalfNormalPrior(scale=0.01),
         )
 
         model = ExactGPModel(X, y, self.likelihood)
@@ -110,8 +111,8 @@ class ExactGPModel(gpytorch.models.ExactGP):
         # q * time
         self.covar_module = (
             (self.cov_stage()
-             * self.cov_time(ls_prior=GammaPrior(concentration=10, rate=5)))
-            + (self.cov_stage()
+             * self.cov_time(ls_prior=GammaPrior(concentration=1,  rate=1)))
+             + (self.cov_stage(ls_prior=GammaPrior(concentration=1, rate=2))
                * self.cov_time(ls_prior=GammaPrior(concentration=2, rate=5))
                * SigmoidKernel(
                    active_dims=self.stage_dim,

rating_gp/providers/usgs.py

@@ -21,6 +21,16 @@ if TYPE_CHECKING:
 FT_TO_M = 0.3048
 FT3_TO_M3 = 0.0283168
 
+    # Quantitative values of "measured_rating_diff"
+USGS_QUALITY_CODES = {
+    'Excellent': '0.02',
+    'Good': '0.05',
+    'Fair': '0.08',
+    'Poor': '0.12',
+    'Unspecified': '0.12',
+}
+
+
 @dataclass
 class NWISColumn:
     column_name: str
@@ -182,6 +192,53 @@ def get_measurements(
     )
     df = nwis._read_rdb(response.text)
 
+    return read_measurements_df(df)    
+
+
+def read_measurements_df(df: pd.DataFrame) -> xr.Dataset:
+    """Read a DataFrame of USGS discharge measurements and convert to xarray Dataset.
+
+    Parameters
+    ----------
+    df : pd.DataFrame
+        Dataframe from `dataretrieval.nwis.get_discharge_measurements()`
+
+    Returns
+    -------
+    xr.Dataset
+    
+    Example
+    -------
+    >>> from dataretrieval import nwis
+    >>> from rating_gp.providers.usgs import read_measurements_df
+    >>> df, _ = nwis.get_discharge_measurements(
+        sites='03339000',
+        start='2020-01-01',
+        end='2020-12-31',
+        format='rdb_expanded',
+        )
+    >>> ds = read_measurements_df(df)
+    """
+
+    # assert the correct columns are present
+    required_columns = [
+        "measurement_dt",
+        "gage_height_va",
+        "discharge_va",
+        "q_meas_used_fg",
+        "control_type_cd",
+        "measured_rating_diff",
+        "streamflow_method",
+        NWISStage.column_name,
+        NWISDischarge.column_name,
+    ]
+
+    missing_columns = set(required_columns) - set(df.columns)
+    if missing_columns:
+        raise ValueError(
+            f"Missing required columns in the DataFrame: {missing_columns}"
+        )
+
     # covert timezone to UTC? ignore for now
     df.index = pd.to_datetime(
         df["measurement_dt"],
@@ -215,26 +272,16 @@ def get_measurements(
             UserWarning,
         )
 
-    
-
-    # Quantitative values of "measured_rating_diff"
-    quality_codes = {
-        'Excellent': '0.02',
-        'Good': '0.05',
-        'Fair': '0.08',
-        'Poor': '0.12',
-        'Unspecified': '0.12',
-    }
     # Replace other values with 'Unspecified'
     df['measured_rating_diff'] = df['measured_rating_diff'].where(
-        df['measured_rating_diff'].isin(quality_codes),
+        df['measured_rating_diff'].isin(USGS_QUALITY_CODES.keys()),
         'Unspecified'
     )
 
-
     df['discharge_unc_frac'] = (df['measured_rating_diff']
-                                .replace(quality_codes)
+                                .replace(USGS_QUALITY_CODES)
                                 .astype(float))
+
     # Set indirect measurements as 20% uncertain regardless of quality code
     df.loc[df['streamflow_method'] == 'QIDIR', 'discharge_unc_frac'] = 0.2