def load_results(folder, quantity, acro_dict):
df = pd.DataFrame()
indices = 0
result_folder_ = f"{folder}/{acro_dict['brick']}/results"
for file in os.listdir(result_folder_):
if quantity in file:
value_dict = delphin_parser.d6o_to_dict(result_folder_,
file)[0]['result']
cell = list(value_dict.keys())[0]
values = value_dict[cell]
df[file.split('.')[0][-1]] = values
indices += 1
df.index = pd.DatetimeIndex(start=datetime.datetime(2019, 1, 1),
freq='h',
periods=len(values))
df.columns = pd.MultiIndex.from_arrays(
[df.columns, [
'brick',
] * indices, [
'out',
] * indices],
names=['location', 'simulation type', 'value type'])
indices = 0
result_folder_ = f"{folder}/{acro_dict['mortar']}/results"
for file in os.listdir(result_folder_):
if quantity in file:
value_dict = delphin_parser.d6o_to_dict(result_folder_,
file)[0]['result']
cell = list(value_dict.keys())[0]
values = value_dict[cell]
df[file.split('.')[0][-1], 'mortar', 'out'] = values
indices += 1
indices = 0
result_folder_ = f"{folder}/{acro_dict['2d']}/results"
for file in os.listdir(result_folder_):
if quantity in file:
value_dict = delphin_parser.d6o_to_dict(result_folder_,
file)[0]['result']
cell = list(value_dict.keys())[0]
values = value_dict[cell]
df[file.split('.')[0][-1], '2d', 'out'] = values
indices += 1
df = df.sort_index(axis=1)
return df
def upload_results_to_database(path_: str, delete_files: bool =True) -> str:
"""
Uploads the results from a Delphin simulation.
:param path_: folder path containing the result files
:param delete_files: if True the result folder will be deleted. Default is True
:return: Result entry id
"""
id_ = os.path.split(path_)[1]
delphin_entry = delphin_db.Delphin.objects(id=id_).first()
result_dict = {}
result_path = path_ + '/results'
log_path = path_ + '/log'
geometry_dict = {}
meta_dict = {}
for result_file in os.listdir(result_path):
if result_file.endswith('.d6o'):
result_dict[result_file.split('.')[0]], meta_dict = delphin_parser.d6o_to_dict(result_path, result_file)
elif result_file.endswith('.g6a'):
geometry_dict = delphin_parser.g6a_to_dict(result_path, result_file)
entry = result_db.Result()
entry.delphin = delphin_entry
entry.log['integrator_cvode_stats'] = delphin_parser.cvode_stats_to_dict(log_path)
entry.log['les_direct_stats'] = delphin_parser.les_stats_to_dict(log_path)
entry.log['progress'] = delphin_parser.progress_to_dict(log_path)
entry.geometry_file = geometry_dict
entry.results = result_dict
entry.simulation_started = meta_dict['created']
entry.geometry_file_hash = meta_dict['geo_file_hash']
entry.save()
# Add results reference to Delphin entry
delphin_entry.update(push__results_raw=entry)
if delete_files:
shutil.rmtree(path_)
return entry.id
map_projects = {
'5ad723b82e2cb22ff0a202f1': 'brick_1d',
'5ad7240c2e2cb22ff0a20333': 'mortar_1d',
'5ad858392e2cb24344ad4ecb': '2d'
}
for project in projects:
for mp_key in map_projects.keys():
if project == mp_key:
key = map_projects[mp_key]
folder = result_folder + f'/{project}/results'
geo_file = [file for file in os.listdir(folder)
if file.endswith('.g6a')][0]
parsed_dicts[key]['moisture'], _ = delphin_parser.d6o_to_dict(
folder, files[0])
parsed_dicts[key]['geo'] = delphin_parser.g6a_to_dict(folder, geo_file)
x = np.linspace(
0, len(parsed_dicts['brick_1d']['moisture']['result']['cell_0'][8760:]),
len(parsed_dicts['brick_1d']['moisture']['result']['cell_0'][8760:]))
x_date = [
datetime.datetime(2020, 1, 1) + datetime.timedelta(hours=i) for i in range(
len(parsed_dicts['brick_1d']['moisture']['result']['cell_0'][8760:]))
]
x_2d = np.linspace(
0, len(parsed_dicts['2d']['moisture']['result']['cell_66'][8760:]),
len(parsed_dicts['2d']['moisture']['result']['cell_66'][8760:]))
x_date_2d = [
datetime.datetime(2020, 1, 1) + datetime.timedelta(hours=i) for i in range(
len(parsed_dicts['2d']['moisture']['result']['cell_66'][8760:]))
def main(project_):
projects = list(project_dict[project_]['map'].keys())
parsed_dicts = {
'brick_1d': {
'rh': {},
'geo': {}
},
'mortar_1d': {
'rh': {},
'geo': {}
},
'2d': {
'rh': {},
'geo': {}
},
}
for p_ in projects:
for mp_key in project_dict[project_]['map'].keys():
if p_ == mp_key:
key = project_dict[project_]['map'][mp_key]
folder = result_folder + f'/{p_}/results'
geo_file = [
file for file in os.listdir(folder) if file.endswith('.g6a')
][0]
parsed_dicts[key]['rh'], _ = delphin_parser.d6o_to_dict(
folder, files[0])
parsed_dicts[key]['geo'] = delphin_parser.g6a_to_dict(folder, geo_file)
x_date = [
datetime.datetime(2020, 1, 1) + datetime.timedelta(hours=i)
for i in range(
len(parsed_dicts['brick_1d']['rh']['result']['cell_0'][8760:]))
]
# Brick 1D
brick_1d = get_points(parsed_dicts['brick_1d']['rh'],
parsed_dicts['brick_1d']['geo'])
brick_1d.sort(key=lambda point: point['x'])
add_data_to_points(brick_1d, parsed_dicts['brick_1d']['rh'],
'relative_humidity')
# Mortar 1D
mortar_1d = get_points(parsed_dicts['mortar_1d']['rh'],
parsed_dicts['mortar_1d']['geo'])
mortar_1d.sort(key=lambda point: point['x'])
add_data_to_points(mortar_1d, parsed_dicts['mortar_1d']['rh'],
'relative_humidity')
# 2D
sim_2d = get_points(parsed_dicts['2d']['rh'], parsed_dicts['2d']['geo'])
sim_2d.sort(key=lambda point: (point['x'], point['y']))
add_data_to_points(sim_2d, parsed_dicts['2d']['rh'], 'relative_humidity')
# Plots
def plot_locations(quantity):
# Axes 00
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[0]['x']:.4f} and 2D-Location: {sim_2d[0]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[0][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[0][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[0][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[1][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[2][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
# Axes 01
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[1]['x']:.4f} and 2D-Location: {sim_2d[3]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[1][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[1][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[3][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[4][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[5][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
# Axes 10
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[2]['x']:.4f} and 2D-Location: {sim_2d[6]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[2][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[2][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[6][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[7][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[8][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
# Axes 11
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[3]['x']:.4f} and 2D-Location: {sim_2d[9]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[3][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[3][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[9][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[10][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[11][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
# Axes 20
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[4]['x']:.4f} and 2D-Location: {sim_2d[12]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[4][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[4][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[12][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[13][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[14][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
# Axes 21
plt.figure()
plt.title(
f"{quantity}\n1D-Location: {brick_1d[5]['x']:.4f} and 2D-Location: {sim_2d[15]['x']:.4f}"
)
plt.plot(x_date,
brick_1d[5][quantity],
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_1d[5][quantity],
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.plot(x_date,
sim_2d[15][quantity],
color=colors['bottom'],
label=f"2D Bottom")
plt.plot(x_date,
sim_2d[16][quantity],
color=colors['mid'],
label=f"2D Mid")
plt.plot(x_date,
sim_2d[17][quantity],
color=colors['top'],
label=f"2D Top")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel(f'{quantity}')
#plot_locations(quantity='relative_humidity')
#plt.show()
def abs_diff(x1, x2):
return x2 - x1
def rel_diff(x1, x2):
return (abs(x2 - x1)) / x2 * 100
def differences(i, plots=False):
avg_2d = np.mean([
sim_2d[i]['relative_humidity'], sim_2d[i + 2]['relative_humidity'],
sim_2d[i + 2]['relative_humidity']
],
axis=0)
brick_abs = abs_diff(brick_1d[i]['relative_humidity'], avg_2d)
mortar_abs = abs_diff(mortar_1d[i]['relative_humidity'], avg_2d)
brick_rel = rel_diff(brick_1d[i]['relative_humidity'], avg_2d)
mortar_rel = rel_diff(mortar_1d[i]['relative_humidity'], avg_2d)
if plots:
# Plot
plt.figure()
plt.title(
f"relative_humidity - Absolute Difference\n"
f"1D-Location: {brick_1d[i]['x']:.4f} and 2D-Location: {sim_2d[i*3]['x']:.4f}"
)
plt.plot(x_date,
brick_abs,
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_abs,
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel('%')
plt.figure()
plt.title(
f"relative_humidity - Relative Difference\n"
f"1D-Location: {brick_1d[i]['x']:.4f} and 2D-Location: {sim_2d[i*3]['x']:.4f}"
)
plt.plot(x_date,
brick_rel,
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_rel,
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel('%')
local_df = pd.DataFrame(
columns=[
f"{brick_1d[i]['x']:.04f}", f"{brick_1d[i]['x']:.04f}",
f"{brick_1d[i]['x']:.04f}", f"{brick_1d[i]['x']:.04f}"
],
index=pd.DatetimeIndex(start=datetime.datetime(2020, 1, 1),
freq='h',
periods=len(brick_rel)),
data=np.vstack([brick_rel, brick_abs, mortar_rel, mortar_abs]).T)
local_df.columns = pd.MultiIndex.from_arrays(
[
local_df.columns, ['brick', 'brick', 'mortar', 'mortar'],
['relative', 'absolute', 'relative', 'absolute']
],
names=['location', 'material', 'type'])
return local_df
def differences_weighted(i, plots=False):
avg_2d = np.average(a=[
sim_2d[i]['relative_humidity'], sim_2d[i + 2]['relative_humidity'],
sim_2d[i + 2]['relative_humidity']
],
axis=0,
weights=[56., 24., 56.])
brick_abs = abs_diff(brick_1d[i]['relative_humidity'], avg_2d)
mortar_abs = abs_diff(mortar_1d[i]['relative_humidity'], avg_2d)
brick_rel = rel_diff(brick_1d[i]['relative_humidity'], avg_2d)
mortar_rel = rel_diff(mortar_1d[i]['relative_humidity'], avg_2d)
if plots:
# Plot
plt.figure()
plt.title(
f"Relative Humidity - Weighted Absolute Difference\n"
f"1D-Location: {brick_1d[i]['x']:.4f} and 2D-Location: {sim_2d[i*3]['x']:.4f}"
)
plt.plot(x_date,
brick_abs,
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_abs,
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel('%')
plt.figure()
plt.title(
f"Relative Humidity - Weighted Relative Difference\n"
f"1D-Location: {brick_1d[i]['x']:.4f} and 2D-Location: {sim_2d[i*3]['x']:.4f}"
)
plt.plot(x_date,
brick_rel,
color=colors['1d_brick'],
label=f"1D Brick")
plt.plot(x_date,
mortar_rel,
color=colors['1d_mortar'],
label=f"1D Mortar")
plt.legend()
plt.gcf().autofmt_xdate()
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%B'))
plt.ylabel('%')
local_df = pd.DataFrame(
columns=[
f"{brick_1d[i]['x']:.04f}", f"{brick_1d[i]['x']:.04f}",
f"{brick_1d[i]['x']:.04f}", f"{brick_1d[i]['x']:.04f}"
],
index=pd.DatetimeIndex(start=datetime.datetime(2020, 1, 1),
freq='h',
periods=len(brick_rel)),
data=np.vstack([brick_rel, brick_abs, mortar_rel, mortar_abs]).T)
local_df.columns = pd.MultiIndex.from_arrays(
[
local_df.columns, ['brick', 'brick', 'mortar', 'mortar'],
['relative', 'absolute', 'relative', 'absolute']
],
names=['location', 'material', 'type'])
return local_df
dataframes = []
weighted_dataframes = []
for index in range(len(brick_1d)):
dataframes.append(differences(index))
weighted_dataframes.append(differences_weighted(index))
# plt.show()
result_dataframe = pd.concat(dataframes, axis=1)
w_result_dataframe = pd.concat(weighted_dataframes, axis=1)
absolute_df = result_dataframe.loc[:, pd.IndexSlice[:, :, 'absolute']]
absolute_df.columns = absolute_df.columns.droplevel(level=2)
relative_df = result_dataframe.loc[:, pd.IndexSlice[:, :, 'relative']]
relative_df.columns = relative_df.columns.droplevel(level=2)
w_absolute_df = w_result_dataframe.loc[:, pd.IndexSlice[:, :, 'absolute']]
w_absolute_df.columns = w_absolute_df.columns.droplevel(level=2)
w_relative_df = w_result_dataframe.loc[:, pd.IndexSlice[:, :, 'relative']]
w_relative_df.columns = w_relative_df.columns.droplevel(level=2)
def boxplots():
plt.figure()
ax = absolute_df.boxplot()
ax.set_ylim(-20, 20)
ax.set_ylabel('Relative Humidity - %')
ax.set_title('Non-Weighted Absolute Differences')
plt.figure()
ax = w_absolute_df.boxplot()
ax.set_ylim(-20, 20)
ax.set_ylabel('Relative Humidity - %')
ax.set_title('Weighted Absolute Differences')
plt.show()
# boxplots()
out_folder = r'C:\Users\ocni\PycharmProjects\delphin_6_automation\data_process\2d_1d\processed_data'
def excel():
writer = pd.ExcelWriter(out_folder + '/relative_humidity.xlsx')
relative_df.describe().to_excel(writer, 'relative')
w_relative_df.describe().to_excel(writer, 'relative_weighted')
absolute_df.describe().to_excel(writer, 'absolute')
w_absolute_df.describe().to_excel(writer, 'absolute_weighted')
writer.save()
# excel()
def save_relative():
hdf_file = out_folder + '/relative_humidity.h5'
w_relative_df.to_hdf(hdf_file, project_, append=True)
save_relative()
'mortar_1d': {'temp': {}, 'rh': {}, 'm_content': {}, 'moisture': {}, 'geo': {}},
'2d': {'temp': {}, 'rh': {}, 'm_content': {}, 'moisture': {}, 'geo': {}}, }
map_projects = {'5ad5da522e2cb21a90397b85': 'brick_1d', '5ad5dac32e2cb21a90397b86': 'mortar_1d',
'5ad5e05d5d9460d762130f93': '2d'}
for project in projects:
for mp_key in map_projects.keys():
if project == mp_key:
key = map_projects[mp_key]
folder = result_folder + f'/{project}/results'
geo_file = [file
for file in os.listdir(folder)
if file.endswith('.g6a')][0]
parsed_dicts[key]['temp'], _ = delphin_parser.d6o_to_dict(folder, files[0])
parsed_dicts[key]['rh'], _ = delphin_parser.d6o_to_dict(folder, files[1])
parsed_dicts[key]['m_content'], _ = delphin_parser.d6o_to_dict(folder, files[2])
parsed_dicts[key]['moisture'], _ = delphin_parser.d6o_to_dict(folder, files[3])
parsed_dicts[key]['geo'] = delphin_parser.g6a_to_dict(folder, geo_file)
x = np.linspace(0, len(parsed_dicts['brick_1d']['temp']['result']['cell_0'][8760:]),
len(parsed_dicts['brick_1d']['temp']['result']['cell_0'][8760:]))
x_date = [datetime.datetime(2020, 1, 1) + datetime.timedelta(hours=i)
for i in range(len(parsed_dicts['brick_1d']['temp']['result']['cell_0'][8760:]))]
x_2d = np.linspace(0, len(parsed_dicts['2d']['temp']['result']['cell_66'][8760:]),
len(parsed_dicts['2d']['temp']['result']['cell_66'][8760:]))
x_date_2d = [datetime.datetime(2020, 1, 1) + datetime.timedelta(hours=i)
for i in range(len(parsed_dicts['2d']['temp']['result']['cell_66'][8760:]))]
# Brick 1D