Sync: Merge v1.3.1 bug fixes into develop (#96)

* Merge in changes from `main` branch (v1.3 -> v1.3.1) into `develop`, notably:

- Feature: Add PyPI integration by @Bartdoekemeijer in #87
- Bug fix: various small corrections by @Bartdoekemeijer in #86

---------

Co-authored-by: paulf81 <paul.fleming@nrel.gov>
Co-authored-by: Rafael M Mudafort <rafael.mudafort@nrel.gov>

.github/workflows/python-publish.yml

@@ -0,0 +1,30 @@
+# This workflows will upload a Python Package using Twine when a release is created
+# For more information see: https://help.github.com/en/actions/language-and-framework-guides/using-python-with-github-actions#publishing-to-package-registries
+
+name: Upload Python Package
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  deploy:
+    if: github.repository_owner == 'NREL'
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: '3.x'
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install setuptools wheel twine
+    - name: Build and publish
+      env:
+        TWINE_USERNAME: ${{ secrets.PYPI_USERNAME }}
+        TWINE_PASSWORD: ${{ secrets.PYPI_PASSWORD }}
+      run: |
+        python setup.py sdist bdist_wheel
+        twine upload dist/*

README.rst

@@ -1,6 +1,6 @@
-==========
+=============================================
 FLORIS-based Analysis for SCADA data (FLASC)
-==========
+=============================================
 
 **Further documentation is available at http://flasc.readthedocs.io/.**
 

flasc/dataframe_operations/dataframe_filtering.py

@@ -133,7 +133,6 @@ def df_mark_turbdata_as_faulty(df, cond, turbine_list, exclude_columns=[]):
         turbine_list = [turbine_list]
 
     for ti in turbine_list:
-        N_init = df_get_no_faulty_measurements(df, ti)
         cols = [s for s in df.columns if s[-4::] == ('_%03d' % ti)
                 and s not in exclude_columns]
         df.loc[cond, cols] = None  # Delete measurements

flasc/dataframe_operations/dataframe_manipulations.py

@@ -698,7 +698,7 @@ def df_reduce_precision(df_in, verbose=False):
     Returns:
         df_out ([pd.DataFrame]): Reduced dataframe
     """
-    df_out = pd.DataFrame()
+    list_out = []
     dtypes = df_in.dtypes
     for ii, c in enumerate(df_in.columns):
         datatype = str(dtypes[c])
@@ -711,42 +711,45 @@ def df_reduce_precision(df_in, verbose=False):
                                df_in[c], equal_nan=True))):
                 unique_values = np.unique(df_in[c])
                 if np.array_equal(unique_values, [0, 1]):
-                    df_out[c] = df_in[c].astype(bool)
+                    var_downsampled = df_in[c].astype(bool)
                 elif np.max(df_in[c]) < np.iinfo(np.int8).max:
-                    df_out[c] = df_in[c].astype(np.int8)
+                    var_downsampled = df_in[c].astype(np.int8)
                 elif np.max(df_in[c]) < np.iinfo(np.int16).max:
-                    df_out[c] = df_in[c].astype(np.int16)
+                    var_downsampled = df_in[c].astype(np.int16)
                 elif np.max(df_in[c]) < np.iinfo(np.int32).max:
-                    df_out[c] = df_in[c].astype(np.int32)
+                    var_downsampled = df_in[c].astype(np.int32)
                 else:
-                    df_out[c] = df_in[c].astype(np.int64)
+                    var_downsampled = df_in[c].astype(np.int64)
             else:  # If not, just simplify as float32
-                df_out[c] = df_in[c].astype(np.float32)
-            max_error = np.max(np.abs(df_out[c]-df_in[c]))
+                var_downsampled = df_in[c].astype(np.float32)
+            max_error = np.max(np.abs(var_downsampled-df_in[c]))
             if verbose:
                 print("Column %s ['%s'] was downsampled to %s."
-                      % (c, datatype, df_out.dtypes[ii]))
+                      % (c, datatype, var_downsampled.dtypes))
                 print( "Max error: ", max_error)
         elif ((datatype == 'int64') or
               (datatype == 'int32') or
               (datatype == 'int')):
-            if all(np.unique(df_in[c]) == [0, 1]):
-                df_out[c] = df_in[c].astype(bool)
+            if np.array_equal(np.unique(df_in[c]), [0, 1]):
+                var_downsampled = df_in[c].astype(bool)
             elif len(np.unique(df_in[c])) < 100:
-                df_out[c] = df_in[c].astype(np.int16)
+                var_downsampled = df_in[c].astype(np.int16)
             else:
-                df_out[c] = df_in[c].astype(np.int32)
-            max_error = np.max(np.abs(df_out[c]-df_in[c]))
+                var_downsampled = df_in[c].astype(np.int32)
+            max_error = np.max(np.abs(var_downsampled-df_in[c]))
             if verbose:
                 print("Column %s ['%s'] was downsampled to %s."
-                      % (c, datatype, df_out.dtypes[ii]))
+                      % (c, datatype, var_downsampled.dtypes))
                 print( "Max error: ", max_error)
         else:
             if verbose:
                 print("Datatype '%s' not recognized. Not downsampling."
                       % datatype)
-            df_out[c] = df_in[c]
+            var_downsampled = df_in[c]
 
+        list_out.append(var_downsampled)
+
+    df_out = pd.concat(list_out, axis=1, ignore_index=False)
     return df_out
 
 

flasc/energy_ratio/energy_ratio_visualization.py

@@ -134,20 +134,29 @@ def plot(
     for ii, df in enumerate(energy_ratios):
         df = df.copy()
 
+        if df.shape[0] < 2:
+            # Do not plot single values
+            continue
+
         # Get x-axis values
         x = np.array(df["wd_bin"], dtype=float)
 
         # Add NaNs to avoid connecting plots over gaps
         dwd = np.min(x[1::] - x[0:-1])
         jumps = np.where(np.diff(x) > dwd * 1.50)[0]
         if len(jumps) > 0:
-            df = df.append(
-                pd.DataFrame(
-                    {
-                        "wd_bin": x[jumps] + dwd / 2.0,
-                        "N_bin": [0] * len(jumps),
-                    }
-                )
+            df = pd.concat(
+                [
+                    df,
+                    pd.DataFrame(
+                        {
+                            "wd_bin": x[jumps] + dwd / 2.0,
+                            "N_bin": [0] * len(jumps),
+                        }
+                    )
+                ],
+                axis=0,
+                ignore_index=False,
             )
             df = df.iloc[np.argsort(df["wd_bin"])].reset_index(drop=True)
             x = np.array(df["wd_bin"], dtype=float)
@@ -176,7 +185,13 @@ def plot(
             )
 
     # Plot the bin count
-    if df_freqs is not None:
+    is_none = False
+    if df_freqs is None:
+        is_none = True
+    elif isinstance(df_freqs, list):
+        is_none = np.any([c is None for c in df_freqs])
+
+    if not is_none:
         for ii, df_freq in enumerate(df_freqs):
             wd_bins = df_freq["wd_bin"].unique()
             n_ws_bins = len(df_freq["ws_bin_edges"].unique())

flasc/energy_ratio/energy_ratio_wd_bias_estimation.py

@@ -408,7 +408,7 @@ def cost_fun(wd_bias):
                                                xtol=0.1, disp=True)
         )
 
-        dran = opt_search_range[1]-opt_search_range[0]
+        dran = opt_search_range[1] - opt_search_range[0]
         x_opt, J_opt, x, J = opt.brute(
             func=cost_fun,
             ranges=[opt_search_range],

flasc/optimization.py

@@ -259,7 +259,11 @@ def match_y_curves_by_offset(yref, ytest, dy_eval=None, angle_wrapping=True):
             ytest_cor = ytest - dy
             y_error = np.abs(yref-ytest_cor)
 
-        J = np.nanmean(y_error**2.0)
+        if np.all(np.isnan(y_error)) | (len(y_error) < 1):
+            J = np.nan
+        else:
+            J = np.nanmean(y_error**2.0)
+
         if np.isnan(J_opt):
             if not np.isnan(J):
                 J_opt = J