APOG_public/acsmnode
0
0
Fork 0
mirror of https://gitea.psi.ch/APOG/acsmnode.git synced 2025-06-24 05:07:57 +02:00
Code Issues Packages Projects Releases Wiki Activity

Replace nonoperation reviewPannel in app/data_flagging_app.py with date range picker. Cached time column to speed up figure update but it doesnot look there was much improvement.

Browse Source
This commit is contained in:
florez_j
2025-06-04 14:03:45 +02:00
parent da6884ef21
commit effd5b865e
2 changed files with 74 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

70
app/data_flagging_app.py
View File

@ -126,6 +126,17 @@ ReviewOpsPannel = dbc.Col([
#]),
],width=12)
DatePickerRange = dbc.Col([
html.H2("Set date range for time series display", style={'font-size': '20px', 'margin-bottom': '10px'}),
dcc.DatePickerRange(
id='date-picker-range',
display_format='YYYY-MM-DD',
start_date_placeholder_text='Start Date',
end_date_placeholder_text='End Date',
minimum_nights=0,
style={'width': '100%'}
)
])
# Initialize Dash app with Bootstrap theme
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
@ -159,11 +170,12 @@ app.layout = dbc.Container([
#dbc.Col([html.Div(id='flag-record', style={'whiteSpace': 'pre-line'})], width=4), #config={'modeBarButtons': True,
#'modeBarButtonsToAdd':['select2d','lasso2d'],
#'modeBarButtonsToRemove': ['zoom', 'pan']}),], width=12)
dbc.Col([
dbc.Col(
[
html.Div([
EnableVisCheckbox,
FlagVisTable,
ReviewOpsPannel,
DatePickerRange,
],
style={'height': '1000px','overflowY': 'auto'}), # Set a fixed height for the div
],
@ -345,19 +357,26 @@ def update_variable_dropdown(instFolderName, fileName, data):
return [{"label": var_name, "value": var_name} for var_name in variableList] , False, variableList
from datetime import datetime
import numpy as np
@app.callback(
Output('timeseries-plot', 'figure'),
Output('memory-output','data'),
Input('instrument-dropdown', 'value'),
Input('file-dropdown', 'value'),
Input('sub-dropdown', 'value'),
Input('date-picker-range', 'start_date'),
Input('date-picker-range', 'end_date'),
Input('memory-output', 'data'),
prevent_initial_call=True
)
def update_figure(instFolderName, fileName, variableList, data):
# Check if any input is None or empty
def update_figure(instFolderName, fileName, variableList, start_date, end_date, data):
fig = go.Figure() # Always define it to avoid UnboundLocalError
if not all([instFolderName, fileName, variableList, data]):
return go.Figure(), dash.no_update # Return an empty figure to prevent crashes
return go.Figure(), dash.no_update
path_to_file = data.get('path_to_uploaded_file')
if not path_to_file:
@ -366,23 +385,54 @@ def update_figure(instFolderName, fileName, variableList, data):
try:
DataOps = hdf5_ops.HDF5DataOpsManager(path_to_file)
DataOps.load_file_obj()
dataset_name = '/'.join([instFolderName, fileName, 'data_table'])
# Get attributes for data table
dataset_name = '/'.join([instFolderName, fileName, 'data_table'])
datetime_var, datetime_var_format = DataOps.infer_datetime_variable(dataset_name)
DataOps.unload_file_obj()
if not isinstance(data.get('time_cache'), dict):
data['time_cache'] = {}
cache_key = f"{path_to_file}|{dataset_name}|{datetime_var}|{datetime_var_format}"
if cache_key in data['time_cache']:
time_column = np.array(data['time_cache'][cache_key])
else:
time_column = DataOps.reformat_datetime_column(
dataset_name, datetime_var, datetime_var_format
)
data['time_cache'][cache_key] = time_column.astype(str).tolist()
# Convert to datetime64, safely handling NaNs or invalid entries
try:
time_column = np.array(time_column, dtype='datetime64[ns]')
except Exception:
# If conversion fails (e.g. mixed formats), fall back to pandas
import pandas as pd
time_column = pd.to_datetime(time_column, errors='coerce').to_numpy()
# Apply mask if date range provided
mask = ~np.isnat(time_column) # Exclude NaT values
if start_date and end_date:
start = np.datetime64(start_date)
end = np.datetime64(end_date)
mask &= (time_column >= start) & (time_column <= end)
fig, channel_names = data_flagging_utils.create_loaded_file_figure(
path_to_file, instFolderName, dataset_name, datetime_var, datetime_var_format, variableList
path_to_file, instFolderName, dataset_name, time_column, variableList, mask=mask
)
data['channel_names'] = channel_names
except Exception as e:
print(f'While processing file {path_to_file}, we got the following exception {e}.')
print(f'Error while processing file {path_to_file}: {e}')
finally:
DataOps.unload_file_obj()
return fig, data
"""@app.callback(
Output('memory-output','data'),
Output('timeseries-plot', 'figure'),

23
app/data_flagging_utils.py
View File

@ -50,9 +50,7 @@ def filter_flags_by_label(flags_dict, label):
if label == 'all' or value['validity'] == label
]
def create_loaded_file_figure(file_path, instFolder, dataset_name, datetime_var, datetime_var_format, variables):
def create_loaded_file_figure(file_path, instFolder, dataset_name, time_column, variables, mask):
DataOpsAPI = h5de.HDF5DataOpsManager(file_path)
@ -70,16 +68,23 @@ def create_loaded_file_figure(file_path, instFolder, dataset_name, datetime_var,
row_heights = [1 for i in range(len(variables))])
traces = []
trace_idx = 1
indices = np.where(mask)[0]
start_idx = indices[0]
end_idx = indices[-1] + 1 # slice is exclusive end
dataset = DataOpsAPI.file_obj[dataset_name]
time_column = DataOpsAPI.reformat_datetime_column(dataset_name,
datetime_var,
datetime_var_format)
#time_column = DataOpsAPI.reformat_datetime_column(dataset_name,
# datetime_var,
# datetime_var_format)
#time_column = dataset[datetime_var][:]
for i in range(0,len(variables)):
x = time_column[start_idx:end_idx]
y = dataset[variables[i]][start_idx:end_idx]
fig.add_trace(go.Scatter(x = time_column,
y = dataset[variables[i]][:],
fig.add_trace(go.Scatter(x = x,
y = y,
mode = 'lines',
name = variables[i]), row=trace_idx, col=1)
fig.update_yaxes(title_text= variables[i], row=trace_idx, col=1)