eliminate some of the test suite warnings (pvlib#906)

* handle sam_data fixture duplicate entries warnings

* pep8

* force numpy within detect_clearsky to avoid pandas indexing deprecation

* martin_ruiz_diffuse runtimewarnings

* losses dtype, pd.datetime

* whats new

Author: wholmgren

docs/sphinx/source/whatsnew/v0.7.2.rst

@@ -34,7 +34,11 @@ Bug fixes
   0.6.4 what's new file. (:issue:`898`)
 * Compatibility with cftime 1.1. (:issue:`895`)
 * Add Python3.8 to Azure Pipelines CI (:issue:`903`)(:pull:`904`)
-
+* Minor implemention changes to avoid runtime and deprecation warnings in
+  :py:func:`~pvlib.clearsky.detect_clearsky`,
+  :py:func:`~pvlib.iam.martin_ruiz_diffuse`,
+  :py:func:`~pvlib.losses.soiling_hsu`,
+  and various test functions.
 
 Documentation
 ~~~~~~~~~~~~~

pvlib/clearsky.py

@@ -712,6 +712,12 @@ def detect_clearsky(measured, clearsky, times, window_length,
     H = hankel(np.arange(intervals_per_window),                   # noqa: N806
                np.arange(intervals_per_window - 1, len(times)))
 
+    # convert pandas input to numpy array, but save knowledge of input state
+    # so we can return a series if that's what was originally provided
+    ispandas = isinstance(measured, pd.Series)
+    measured = np.asarray(measured)
+    clearsky = np.asarray(clearsky)
+
     # calculate measurement statistics
     meas_mean = np.mean(measured[H], axis=0)
     meas_max = np.max(measured[H], axis=0)
@@ -771,7 +777,7 @@ def rmse(alpha):
                       % max_iterations, RuntimeWarning)
 
     # be polite about returning the same type as was input
-    if isinstance(measured, pd.Series):
+    if ispandas:
         clear_samples = pd.Series(clear_samples, index=times)
 
     if return_components:

pvlib/iam.py

@@ -363,11 +363,13 @@ def martin_ruiz_diffuse(surface_tilt, a_r=0.16, c1=0.4244, c2=None):
 
     from numpy import pi, sin, cos, exp
 
-    # because sin(pi) isn't exactly zero
-    sin_beta = np.where(surface_tilt < 90, sin(beta), sin(pi - beta))
+    # avoid RuntimeWarnings for <, sin, and cos with nan
+    with np.errstate(invalid='ignore'):
+        # because sin(pi) isn't exactly zero
+        sin_beta = np.where(surface_tilt < 90, sin(beta), sin(pi - beta))
 
-    trig_term_sky = sin_beta + (pi - beta - sin_beta) / (1 + cos(beta))
-    trig_term_gnd = sin_beta +      (beta - sin_beta) / (1 - cos(beta)) # noqa: E222 E261 E501
+        trig_term_sky = sin_beta + (pi - beta - sin_beta) / (1 + cos(beta))
+        trig_term_gnd = sin_beta +      (beta - sin_beta) / (1 - cos(beta))  # noqa: E222 E261 E501
 
     iam_sky = 1 - exp(-(c1 + c2 * trig_term_sky) * trig_term_sky / a_r)
     iam_gnd = 1 - exp(-(c1 + c2 * trig_term_gnd) * trig_term_gnd / a_r)

pvlib/losses.py

@@ -78,7 +78,8 @@ def soiling_hsu(rainfall, cleaning_threshold, tilt, pm2_5, pm10,
     tms_cumsum = np.cumsum(tilted_mass_rate * np.ones(rainfall.shape))
 
     mass_no_cleaning = pd.Series(index=rainfall.index, data=tms_cumsum)
-    mass_removed = pd.Series(index=rainfall.index)
+    # specify dtype so pandas doesn't assume object
+    mass_removed = pd.Series(index=rainfall.index, dtype='float64')
     mass_removed[0] = 0.
     mass_removed[cleaning_times] = mass_no_cleaning[cleaning_times]
     accum_mass = mass_no_cleaning - mass_removed.ffill()

pvlib/pvsystem.py

@@ -1546,7 +1546,7 @@ def _normalize_sam_product_names(names):
 
     import warnings
 
-    BAD_CHARS  = ' -.()[]:+/",'
+    BAD_CHARS = ' -.()[]:+/",'
     GOOD_CHARS = '____________'
 
     mapping = str.maketrans(BAD_CHARS, GOOD_CHARS)
@@ -1559,7 +1559,8 @@ def _normalize_sam_product_names(names):
 
     n_duplicates = norm_names.duplicated().sum()
     if n_duplicates > 0:
-        warnings.warn('Normalized names contain %d duplicate(s).' % n_duplicates)
+        warnings.warn(
+            'Normalized names contain %d duplicate(s).' % n_duplicates)
 
     return norm_names.values
 

pvlib/tests/conftest.py

@@ -1,5 +1,6 @@
 from pathlib import Path
 import platform
+import warnings
 
 import numpy as np
 import pandas as pd
@@ -153,8 +154,11 @@ def has_numba():
 @pytest.fixture(scope="session")
 def sam_data():
     data = {}
-    data['sandiamod'] = pvlib.pvsystem.retrieve_sam('sandiamod')
-    data['adrinverter'] = pvlib.pvsystem.retrieve_sam('adrinverter')
+    with warnings.catch_warnings():
+        # ignore messages about duplicate entries in the databases.
+        warnings.simplefilter("ignore", UserWarning)
+        data['sandiamod'] = pvlib.pvsystem.retrieve_sam('sandiamod')
+        data['adrinverter'] = pvlib.pvsystem.retrieve_sam('adrinverter')
     return data
 
 

pvlib/tests/test_losses.py

@@ -15,8 +15,8 @@
 @pytest.fixture
 def expected_output():
     # Sample output (calculated manually)
-    dt = pd.date_range(start=pd.datetime(2019, 1, 1, 0, 0, 0),
-                       end=pd.datetime(2019, 1, 1, 23, 59, 0), freq='1h')
+    dt = pd.date_range(start=pd.Timestamp(2019, 1, 1, 0, 0, 0),
+                       end=pd.Timestamp(2019, 1, 1, 23, 59, 0), freq='1h')
 
     expected_no_cleaning = pd.Series(
         data=[0.884980357535360, 0.806308930084762, 0.749974647038078,
@@ -35,12 +35,12 @@ def expected_output():
 @pytest.fixture
 def expected_output_2(expected_output):
     # Sample output (calculated manually)
-    dt = pd.date_range(start=pd.datetime(2019, 1, 1, 0, 0, 0),
-                       end=pd.datetime(2019, 1, 1, 23, 59, 0), freq='1h')
+    dt = pd.date_range(start=pd.Timestamp(2019, 1, 1, 0, 0, 0),
+                       end=pd.Timestamp(2019, 1, 1, 23, 59, 0), freq='1h')
 
     expected_no_cleaning = expected_output
 
-    expected = pd.Series(index=dt)
+    expected = pd.Series(index=dt, dtype='float64')
     expected[dt[:4]] = expected_no_cleaning[dt[:4]]
     expected[dt[4:7]] = 1.
     expected[dt[7]] = expected_no_cleaning[dt[0]]
@@ -55,8 +55,8 @@ def expected_output_2(expected_output):
 @pytest.fixture
 def rainfall_input():
 
-    dt = pd.date_range(start=pd.datetime(2019, 1, 1, 0, 0, 0),
-                       end=pd.datetime(2019, 1, 1, 23, 59, 0), freq='1h')
+    dt = pd.date_range(start=pd.Timestamp(2019, 1, 1, 0, 0, 0),
+                       end=pd.Timestamp(2019, 1, 1, 23, 59, 0), freq='1h')
     rainfall = pd.Series(
         data=[0., 0., 0., 0., 1., 0., 0., 0., 0.5, 0.5, 0., 0., 0., 0., 0.,
               0., 0.3, 0.3, 0.3, 0.3, 0., 0., 0., 0.], index=dt)

pvlib/tests/test_pvsystem.py

@@ -1085,7 +1085,7 @@ def test_singlediode_series_ivcurve(cec_module_params):
         assert_allclose(v, expected[k], atol=1e-2)
 
 
-def test_scale_voltage_current_power(sam_data):
+def test_scale_voltage_current_power():
     data = pd.DataFrame(
         np.array([[2, 1.5, 10, 8, 12, 0.5, 1.5]]),
         columns=['i_sc', 'i_mp', 'v_oc', 'v_mp', 'p_mp', 'i_x', 'i_xx'],