Move map setup to separate module and add run buttons

notebooks/map.py

@@ -0,0 +1,274 @@
+import os
+import shapely.geometry
+import numpy as np
+from birdy import WPSClient
+from netCDF4 import Dataset
+from datetime import datetime
+from requests_html import HTMLSession
+from ipywidgets import *
+from ipyleaflet import *
+from IPython import display as ipydisplay
+
+# Instantiate the clients to the two birds. This instantiation also takes advantage of asynchronous execution by setting `progress` to True.
+chickadee_url = "http://docker-dev03.pcic.uvic.ca:30102"
+chickadee = WPSClient(chickadee_url, progress=True)
+finch_url = "https://docker-dev03.pcic.uvic.ca/twitcher/ows/proxy/finch/wps"
+finch = WPSClient(finch_url, progress=True)
+
+# These outputs store the WPS responses to track the bird processes
+chickadee_output = None
+finch_output = None
+
+sub_layers = LayerGroup()
+thredds_base = "https://docker-dev03.pcic.uvic.ca/twitcher/ows/proxy/thredds/dodsC/datasets"
+thredds_catalog = "https://docker-dev03.pcic.uvic.ca/twitcher/ows/proxy/thredds/catalog/datasets"
+
+def in_bc(point):
+    bc = f"{thredds_base}/storage/data/climate/PRISM/dataportal/pr_monClim_PRISM_historical_run1_197101-200012.nc"
+    bc_data = Dataset(bc)
+    bc_lat = bc_data.variables["lat"][:]
+    bc_lon = bc_data.variables["lon"][:]
+    # Check if center point is within lat/lon grid
+    if (point[0] < bc_lat[0]) or (point[0] > bc_lat[-1]) or (point[1] < bc_lon[0]) or (point[1] > bc_lon[-1]):
+        return False
+    # Check if center point is closest to a masked data value
+    else:
+        lat_index = np.argmin(np.abs(bc_lat - point[0]))
+        lon_index = np.argmin(np.abs(bc_lon - point[1]))
+        pr = bc_data.variables["pr"][0, lat_index, lon_index]
+        if pr.mask:
+            return False
+    return True
+
+def get_subdomain(lat_min, lat_max, lon_min, lon_max, color, name):
+    coords = [(lat_min, lon_min), (lat_max, lon_max)]
+    return Rectangle(bounds=coords, color=color, name=name, draggable=True)
+
+def get_models():
+    session = HTMLSession()
+    r = session.get(f"{thredds_catalog}/storage/data/climate/downscale/BCCAQ2/CMIP6_BCCAQv2/catalog.html")
+    models = []
+    exclude = ["CMIP6_BCCAQv2", "CWEC2020_Factors/", "Degree_Climatologies/", "Ensemble_Averages/", "nobackup/", "--", ""]
+    for tt in r.html.find("tt"):
+        if tt.text not in exclude:
+            models.append(tt.text[:-1])
+    models.sort()
+    return models
+
+def handle_dataset_change(change):
+    technique.disabled = not technique.disabled
+    model.disabled = not model.disabled
+    scenario.disabled = not scenario.disabled
+
+def handle_model_change(change):
+    if model.value == "CanESM5":
+        canesm5_run.disabled = False
+    else:
+        canesm5_run.disabled = True
+
+def handle_interact(**kwargs):
+    point = (round(kwargs.get("coordinates")[0], 5), round(kwargs.get("coordinates")[1], 5))
+    center_hover.value = str(point)
+    if kwargs.get("type") == "click":
+        # Check if point is within PRISM region
+        if not in_bc(point):
+            return
+        # Remove previous center point and subdomains
+        for layer in sub_layers.layers:
+            sub_layers.remove_layer(layer)
+
+        # Add new subdomains
+        m.center_point = point
+        center.value = str(m.center_point)
+        center_marker = Marker(location=m.center_point, name="Marker")
+
+        m.lat_min_obs, m.lat_max_obs = (m.center_point[0] - 1.25, m.center_point[0] + 1.25)
+        m.lon_min_obs, m.lon_max_obs = (m.center_point[1] - 1.25, m.center_point[1] + 1.25)
+        m.lat_min_gcm, m.lat_max_gcm = (m.center_point[0] - 1.5, m.center_point[0] + 1.5)
+        m.lon_min_gcm, m.lon_max_gcm = (m.center_point[1] - 1.5, m.center_point[1] + 1.5)
+
+        gcm_subdomain = get_subdomain(m.lat_min_gcm, m.lat_max_gcm, m.lon_min_gcm, m.lon_max_gcm, "blue", "GCM")
+        obs_subdomain = get_subdomain(m.lat_min_obs, m.lat_max_obs, m.lon_min_obs, m.lon_max_obs, "red", "Obs")
+
+        sub_layers.add_layer(center_marker)
+        sub_layers.add_layer(gcm_subdomain)
+        sub_layers.add_layer(obs_subdomain)
+        m.add_layer(sub_layers)
+
+output_widget = Output() # Used to print bird progress to main workflow
+@output_widget.capture()
+def handle_run_chickadee(arg):
+    # Obtain the input data files from the THREDDS data server (https://docker-dev03.pcic.uvic.ca/twitcher/ows/proxy/thredds/catalog.html)
+    data_vars = {"pr": "pr", "tasmax": "tmax", "tasmin": "tmin"}
+
+    gcm_var = clim_vars.value
+    obs_var = data_vars[gcm_var]
+
+    if dataset.value == "PNWNAmet":
+        gcm_file = f"{thredds_base}/storage/data/projects/dataportal/data/vic-gen2-forcing/PNWNAmet_{gcm_var}_invert_lat.nc"
+    else:
+        if technique.value == "BCCAQv2":
+            technique_dir = "BCCAQ2"
+            model_dir = model.value
+        else:
+            technique_dir = "MBCn"
+            model_dir = model.value + "_10"
+        model_catalog = f"{thredds_catalog}/storage/data/climate/downscale/{technique_dir}/CMIP6_{technique.value}/{model_dir}/catalog.html"
+
+        session = HTMLSession()
+        r = session.get(model_catalog)
+        for tt in r.html.find("tt"):
+            file = tt.text
+            if (gcm_var in file) and (scenario.value in file):
+                if (model.value == "CanESM5") and (canesm5_run.value not in file):
+                    continue
+                break
+        gcm_file = f"{thredds_base}/storage/data/climate/downscale/{technique_dir}/CMIP6_{technique.value}/{model_dir}/{file}"
+
+    obs_file = f"{thredds_base}/storage/data/climate/PRISM/dataportal/{obs_var}_monClim_PRISM_historical_run1_{period.value}.nc"
+    gcm_dataset = Dataset(gcm_file)
+    obs_dataset = Dataset(obs_file)
+
+    # Obtain the datasets' latitudes and longitudes to determine the subdomains
+    gcm_lats = gcm_dataset.variables["lat"][:]
+    gcm_lons = gcm_dataset.variables["lon"][:]
+    obs_lats = obs_dataset.variables["lat"][:]
+    obs_lons = obs_dataset.variables["lon"][:]
+    gcm_lat_indices = get_index_range(gcm_lats, m.lat_min_gcm, m.lat_max_gcm)
+    gcm_lon_indices = get_index_range(gcm_lons, m.lon_min_gcm, m.lon_max_gcm)
+    obs_lat_indices = get_index_range(obs_lats, m.lat_min_obs, m.lat_max_obs)
+    obs_lon_indices = get_index_range(obs_lons, m.lon_min_obs, m.lon_max_obs)
+    gcm_lat_range = f"[{gcm_lat_indices[0]}:{gcm_lat_indices[1]}]"
+    gcm_lon_range = f"[{gcm_lon_indices[0]}:{gcm_lon_indices[1]}]"
+    obs_lat_range = f"[{obs_lat_indices[0]}:{obs_lat_indices[1]}]"
+    obs_lon_range = f"[{obs_lon_indices[0]}:{obs_lon_indices[1]}]"
+
+    # Use full time range of each dataset
+    gcm_ntime = len(gcm_dataset.variables["time"][:])
+    gcm_time_range = f"[0:{gcm_ntime - 1}]"
+    obs_ntime = len(obs_dataset.variables["time"][:])
+    obs_time_range = f"[0:{obs_ntime - 1}]"
+    gcm_dataset.close()
+    obs_dataset.close()
+
+    # Request a subset of each dataset based on the array indices for each subdomain
+    gcm_subset_file = f"{gcm_file}?time,lat{gcm_lat_range},lon{gcm_lon_range},{gcm_var}{gcm_time_range}{gcm_lat_range}{gcm_lon_range}"
+    obs_subset_file = f"{obs_file}?time,lat{obs_lat_range},lon{obs_lon_range},climatology_bounds,crs,{obs_var}{obs_time_range}{obs_lat_range}{obs_lon_range}"
+
+    # Put together the parameters for `chickadee.ci`. 
+    # In the case for `pr`, the `units_bool` parameter is set to `False` in order to avoid converting the PRISM's `mm` units to the PNWNAmet's `mm/day` units.
+    (start_date, end_date) = period.value.split("-")
+    start_date = datetime.strptime(start_date, "%Y%m")
+    end_date = datetime.strptime(end_date + "31", "%Y%m%d")
+    chickadee_params = {"gcm_file": gcm_subset_file, "obs_file": obs_subset_file, "gcm_varname": gcm_var, "obs_varname": obs_var, "max_gb": 0.5, 
+              "start_date": start_date, "end_date": end_date}
+    if gcm_var == "pr":
+        chickadee_params["units_bool"] = False
+        chickadee_params["pr_units"] = "mm/day"
+    if dataset.value == "PNWNAmet":
+        chickadee_params["out_file"] = f"{gcm_var}_{dataset.value}_target_{period.value}_on-demand.nc"
+    else:
+        if not canesm5_run.disabled:
+            chickadee_params["out_file"] = f"{gcm_var}_{dataset.value}_{technique.value}_{model.value}_{canesm5_run.value}_{scenario.value}_target_{period.value}_on-demand.nc"
+        else:
+            chickadee_params["out_file"] = f"{gcm_var}_{dataset.value}_{technique.value}_{model.value}_{scenario.value}_target_{period.value}_on-demand.nc"
+
+    print("Running Chickadee")
+    global chickadee_output
+    chickadee_output = chickadee.ci(**chickadee_params)
+
+@output_widget.capture()
+def handle_run_finch(arg):
+    if chickadee_output is None or chickadee_output.isNotComplete():
+        print("Chickadee process has not completed. Cannot run Finch without the Chickadee output.")
+        return
+
+    # Get THREDDS location of large Chickadee output to pass to Finch
+    chickadee_output_url = chickadee_output.get()[0]
+    chickadee_output_thredds = thredds_base + "birdhouse_wps_outputs/"
+    chickadee_output_thredds += chickadee_output_url.split("wpsoutputs/")[1]
+
+    # Get climate index corresponding to climate variable in Chickadee output
+    with Dataset(chickadee_output_thredds) as dset:
+        if "pr" in dset.variables:
+            process = finch.wetdays
+            kwargs = {"thresh": "10 mm/day", "output_name": "r10mm"}
+        elif "tasmin" in dset.variables:
+            process = finch.tn_min
+            kwargs = {"output_name": "tn_min_ys"}
+        else:
+            process = finch.tx_max
+            kwargs = {"output_name": "tx_max_ys"}
+
+    print("Running Finch")
+    global finch_output
+    finch_output = process(chickadee_output_thredds, **kwargs)
+
+def get_index_range(arr, min_val, max_val):
+    """Compute the indices in an array that correspond to the array's values
+    closest to desired min/max values."""        
+    min_index = np.argmin(np.abs(arr - min_val))
+    max_index = np.argmin(np.abs(arr - max_val))
+    return (min_index, max_index)
+
+def get_bird_output(resp):
+    """Get the URL of the Chickaee/Finch output file for downloading."""
+    if resp.isNotComplete():
+        print("Bird process is not complete.")
+    else:
+        print("Process status: " + resp.status)
+        print("Link to process output: " + resp.get()[0])
+
+def bird_output_to_dataset(resp):
+    """Open Chickadee/Finch output via its THREDDS location using netCDF4.Dataset for further examination."""
+    url = resp.get()[0]
+    thredds_url = thredds_base + "birdhouse_wps_outputs/" + url.split("wpsoutputs")[1]
+    ds = Dataset(thredds_url)
+    return ds
+
+# Initialize interactive map and associated widgets
+mapnik = basemap_to_tiles(basemaps.OpenStreetMap.Mapnik)
+mapnik.base = True
+mapnik.name = "Default"
+
+m = Map(
+    basemap=mapnik,
+    center=(53.5, -120),
+    zoom=5,
+    layout=Layout(height="600px"),
+)
+m.on_interaction(handle_interact)
+m.center_point = ()
+
+legend = LegendControl({"GCM": "blue", "Obs": "red"}, name="Subdomains", position="topright")
+m.add_control(legend)
+
+center_hover = Text(value="", placeholder="") 
+center = Text(value="", placeholder="", description="Center:")
+clim_vars = RadioButtons(options=["pr", "tasmax", "tasmin"], description="Climate variable:")
+dataset = RadioButtons(options=["PNWNAmet", "CMIP6"], description="Dataset:")
+dataset.observe(handle_dataset_change)
+
+technique = RadioButtons(options=["BCCAQv2", "MBCn"], description="CMIP6 downscaling technique:", disabled = True)
+model = Dropdown(options=get_models(), description="CMIP6 model:", disabled = True)
+model.style.description_width = "100px"
+model.observe(handle_model_change)
+
+canesm5_runs = ["r" + str(r) + "i1p2f1" for r in range(1, 11)]
+canesm5_run = Dropdown(options=canesm5_runs, description="CanESM5 run:", disabled = True)
+canesm5_run.style.description_width = "100px"
+
+scenario = RadioButtons(options=[("SSP1-2.6", "ssp126"), ("SSP2-4.5", "ssp245"), ("SSP5-8.5", "ssp585")], description="CMIP6 emissions scenario:", disabled = True)
+period = RadioButtons(options=["197101-200012", "198101-201012"], description="Climatological period:")
+
+run_chickadee = Button(description='Run Chickadee', button_style='success', disabled=False, tooltip="Click 'Run' to start the on-demand downscaling")
+run_chickadee.on_click(handle_run_chickadee)
+
+run_finch = Button(description='Run Finch', button_style='info', disabled=False, tooltip="Click 'Run' to start the climate index calculations")
+run_finch.on_click(handle_run_finch)
+
+box_layout = Layout(display='flex',
+                flex_flow = 'column', 
+                width='110%',
+                align_items = 'center')
+control_box = Box(children = [center_hover, center, clim_vars, dataset, technique, model, canesm5_run, scenario, period, run_chickadee, run_finch], layout=box_layout)