Updated connectivity based on grid, removed the cancellation of Slage…

…lse-Nyborg line, since it makes the clustering algorithm not work which may explain weird connections, and added plotting of line capacities for clustered data

src/Modules/aggregate_inputs.py

@@ -10,11 +10,13 @@
 ###        0. Script Settings       ###
 ### ------------------------------- ###
 
+import matplotlib.pyplot as plt 
 import os 
 import pandas as pd
 import numpy as np
 import geopandas as gpd
 import click
+from typing import Union
 from pybalmorel import Balmorel, IncFile
 from pybalmorel.utils import symbol_to_df
 import gams
@@ -175,6 +177,10 @@ def aggregate_parameter(db: gams.GamsDatabase,
                 prefix=prefix, suffix=suffix)
     f.body = df
 
+    # Plot connection costs in case it is a connection parameter
+    if 'INVCOST' in symbol:
+        plot_transmission_invcost(symbol, df)
+
     # Use N-1 sets as index, and the last as columns, where N = length of columns without 'Value' column
     if len(symbol_columns) > 2:
         index = symbol_columns[:-2]
@@ -223,6 +229,35 @@ def aggregate_sets(db: gams.GamsDatabase,
 
     f.save()
 
+def plot_transmission_invcost(symbol: str,
+                              df: pd.DataFrame,
+                              year: int = 2050):
+    print('Plotting connection costs for %s'%symbol)
+
+    # f = gpd.read_file('ClusterOutput/%s_%dcluster_geofile.gpkg'%('-'.join(cluster_params_list), cluster_size))
+    geo_union = gpd.read_file('ClusterOutput/DE-DH-WNDFLH-SOLEFLH_30cluster_geofile.gpkg')
+    geo_union['coord'] = geo_union.centroid
+
+    fig, ax = plt.subplots()
+    geo_union.plot(ax=ax)
+
+    # lines = []
+    exclusion = []
+    for line, cap in df.loc[str(year)].iterrows():
+        if line in exclusion or line[0] == line[1]:
+            continue
+
+        coords = geo_union.query('cluster_name == @line[0] or cluster_name == @line[1]')['coord']
+        x1, x2 = coords.x
+        y1, y2 = coords.y
+
+        l, = ax.plot([x1,x2], [y1,y2], color = 'r', linewidth = cap['Value']*1e-5)
+        # lines.append(l)
+
+        exclusion.append((line[1], line[0]))
+    # print(df.drop)
+    plt.show()
+
 
 #%% ------------------------------- ###
 ###            2. Main              ###
@@ -231,13 +266,15 @@ def aggregate_sets(db: gams.GamsDatabase,
 @click.command()
 @click.option('--model-path', type=str, required=True, help='Balmorel model path')
 @click.option('--scenario', type=str, required=True, help='Balmorel scenario')
-@click.option('--exceptions', type=str, required=False, help='.inc files that should NOT be generated')
-@click.option('--mean-aggfuncs', type=str, required=False, help='Parameters that should be aggregated with an average')
-@click.option('--median-aggfuncs', type=str, required=False, help='Parameters that should be aggregated using the median value')
-@click.option('--zero-fillnas', type=str, required=False, help='NaN values that should be converted to zero instead of EPS')
-def main(model_path: str, scenario: str, exceptions: str = '', 
-         mean_aggfuncs: str = '', median_aggfuncs: str = '', 
-         zero_fillnas: str = '', incfile_folder: str = 'Output'):
+@click.option('--exceptions', type=str, required=False, default='', help='.inc files that should NOT be generated')
+@click.option('--mean-aggfuncs', type=str, required=False, default='', help='Parameters that should be aggregated with an average')
+@click.option('--median-aggfuncs', type=str, required=False, default='', help='Parameters that should be aggregated using the median value')
+@click.option('--zero-fillnas', type=str, required=False, default='', help='NaN values that should be converted to zero instead of EPS')
+@click.option('--only-symbols', type=str, required=False, help="Only aggregate the symbols, input as comma-separated string")
+def main(model_path: str, scenario: str, exceptions: str, 
+         mean_aggfuncs: str, median_aggfuncs: str, 
+         zero_fillnas: str, incfile_folder: str = 'Output',
+         only_symbols: Union[str, None] = None):
 
     # Make configuration lists
     exceptions = exceptions.replace(' ', '').split(',') # Symbols not to aggregate
@@ -258,6 +295,8 @@ def main(model_path: str, scenario: str, exceptions: str = '',
 
     # Get which .inc-files and how to aggregate based on folder content and configurations
     symbols, aggfuncs, fillnas = get_symbols_to_aggregate(incfile_folder, exceptions, mean_aggfuncs, median_aggfuncs, zero_fillnas)
+    if only_symbols != None:
+        symbols = only_symbols.replace(' ', '').split(',')
 
     # Aggregating parameters and sets
     print('Will attempt to aggregate..\n%s\n'%(','.join(symbols)))

src/Modules/clustering.py

@@ -200,7 +200,6 @@ def cluster(collected_data: pd.DataFrame,
             use_connectivity: bool = True,
             manual_corrections: list = [
                 ['Roedovre', 'Frederiksberg', 1],
-                ['Slagelse', 'Nyborg', 0],
             ],
             linkage: str = 'Ward',
             connection_remark: str = 'connec. included + artifical',
@@ -219,7 +218,7 @@ def cluster(collected_data: pd.DataFrame,
             connectivity.connection.loc[manual_connection[0], manual_connection[1]] = manual_connection[2]
             connectivity.connection.loc[manual_connection[1], manual_connection[0]] = manual_connection[2]
 
-
+        print('Is matrix symmetric?', np.all(connectivity.connection.data == connectivity.connection.data.T))
         ## Make symmetric index, so the indices fit
         collected_data = collected_data.assign_coords(IRRRI=collected_data.coords['IRRRE'].data)
 
@@ -229,6 +228,7 @@ def cluster(collected_data: pd.DataFrame,
 
         ## Make symmetric connectivity graph 
         knn_graph = X.connection.data # get numpy array
+        print('Is matrix still symmetric?', np.all(knn_graph == knn_graph.T))
         knn_graph = csr_matrix(knn_graph) # make dense format
 
         ## Drop the connection variable again
@@ -272,7 +272,7 @@ def cluster(collected_data: pd.DataFrame,
     clustering.plot(ax=ax, column='cluster_group',
                     cmap=truncate_colormap(cmap, 0.2, 1))
 
-    if knn_graph == None:
+    if knn_graph is None:
         connection_remark = 'no connectivity'
 
     plot_title = '%s, %d clusters, %s\ndata: %s'%(linkage,