Dev imrwb

HSP2_DriverInMem.py

@@ -0,0 +1,17 @@
+# sample script to illustrate running HSP2 with inMem version
+
+uciName = "C://dev//HSPsquared//tests//test10//HSP2results//test10.uci"
+wdmName = "C://dev//HSPsquared//tests//test10//HSP2results//test10.wdm"
+
+from readers.utilities.readUCIinMem import readUCI
+from readers.utilities.readWDMinMem import readWDM
+
+uciDict = readUCI(uciName,'')
+wdmDict = readWDM(wdmName,'')
+# combine the 2 dicts
+uciDict.update(wdmDict)
+
+from utilities.HSP2mainInMem import main
+main(uciDict, saveall=True, jupyterlab=False)
+
+x = uciDict

HSP2_DriverInMemUCIonly.py

@@ -0,0 +1,36 @@
+# sample script to illustrate running HSP2 with inMem version
+
+# reads UCI and WDMs into dictionary of pandas dataframes,
+# executes HSP2 off the in-memory dictionary of pandas dataframes,
+# writes output to same dictionary of pandas dataframes
+
+uciName = "C://dev//HSPsquared//tests//test10//HSP2results//test10.uci"
+
+from hsp2.hsp2tools.readUCIinMem import readUCI
+from hsp2.hsp2tools.readWDMinMem import readWDM
+import os
+
+uciDict = readUCI(uciName,'')
+uci_path = os.path.dirname(uciName)
+
+input_file_missing = False
+filesDf = uciDict['/CONTROL/FILES']
+for index, row in filesDf.iterrows():
+    if row['TYPE'][0:3] == 'WDM':
+        wdmName = row['NAME']
+        wdmName = os.path.join(uci_path, wdmName)
+
+        if os.path.isfile(wdmName):
+            wdmDict = readWDM(wdmName,'')
+            # combine the 2 dicts
+            uciDict.update(wdmDict)
+        else:
+            input_file_missing = True
+
+if input_file_missing:
+    print("Cant run " + uciName + ", wdm file missing")
+else:
+    from hsp2.hsp2.HSP2mainInMem import main
+    main(uciDict, saveall=True, jupyterlab=False)
+
+x = uciDict

src/hsp2/hsp2/HSP2utilitiesInMem.py

@@ -0,0 +1,67 @@
+# in memory versions of HSP2 utilities get_timeseries and save_timeseries
+
+from typing import Protocol, Dict, Any, List, Union, runtime_checkable
+from numba import types
+from numba.typed import Dict
+from enum import Enum
+from hsp2.hsp2.utilities import transform, clean_name
+import numpy as np
+import pandas as pd
+
+def get_timeseries_InMem(timeseries_inputs, ext_sourcesdd, siminfo):
+    ''' makes timeseries for the current timestep and trucated to the sim interval'''
+    # explicit creation of Numba dictionary with signatures
+    ts = Dict.empty(key_type=types.unicode_type, value_type=types.float64[:])
+    for row in ext_sourcesdd:
+        path = f'/TIMESERIES/{row.SVOLNO}'
+        if path in timeseries_inputs:
+            data_frame = timeseries_inputs[path]
+            temp_data_frame = data_frame.copy()
+
+        if row.MFACTOR != 1.0:
+            temp_data_frame *= row.MFACTOR
+        t = transform(temp_data_frame, row.TMEMN, row.TRAN, siminfo)
+
+        tname = clean_name(row.TMEMN,row.TMEMSB)
+        if tname in ts:
+            ts[tname] += t
+        else:
+            ts[tname]  = t
+    return ts
+
+def save_timeseries_InMem(timeseries, ts, savedict, siminfo, saveall, operation, segment, activity, compress=True, outstep=2):
+    df = pd.DataFrame(index=siminfo['tindex'])
+    if (operation == 'IMPLND' and activity == 'IQUAL') or (operation == 'PERLND' and activity == 'PQUAL'):
+        for y in savedict.keys():
+            for z in set(ts.keys()):
+                if '/' + y in z:
+                    zrep = z.replace('/','_')
+                    zrep2 = zrep.replace(' ', '')
+                    df[zrep2] = ts[z]
+                if '_' + y in z:
+                    df[z] = ts[z]
+    elif (operation == 'RCHRES' and (activity == 'CONS' or activity == 'GQUAL')):
+        for y in savedict.keys():
+            for z in set(ts.keys()):
+                if '_' + y in z:
+                    df[z] = ts[z]
+        for y in (savedict.keys() & set(ts.keys())):
+            df[y] = ts[y]
+    else:
+        for y in (savedict.keys() & set(ts.keys())):
+            df[y] = ts[y]
+    df = df.astype(np.float32).sort_index(axis='columns')
+
+    if saveall:
+        save_columns = df.columns
+    else:
+        save_columns = [key for key,value in savedict.items() if value or saveall]
+
+    if not df.empty:
+        path = f'{operation}_{segment}/{activity}'
+        #if category:
+        path = '/RESULTS/' + path
+        timeseries[path] = df
+    else:
+        print(f'DataFrame Empty for {operation}|{activity}|{segment}')
+    return