23 map timeline controls

dashboard/bokeh_dash_app.py

@@ -1,11 +1,16 @@
 # load relevant libraries
+import logging
 from bokeh.io import curdoc
 from bokeh.models import TabPanel, Tabs
 
 # load custom modules
 from utilities.bokeh_line_dash import bokeh_line_dash
 from utilities.bokeh_map_dash import bokeh_map_dash
 
+# set up logging
+lgr = logging.getLogger()
+lgr.setLevel(logging.INFO)
+
 # initialise and structure bokeh line dashboard
 dashboard_line = bokeh_line_dash()
 panel_line_tab = TabPanel(child = dashboard_line, title = 'Time Series')

dashboard/cons.py

@@ -61,6 +61,7 @@
 linedash_year_timespan = ["2010", linedash_yearend]
 linedash_yearmonth_timespan = ["2010-01", f"{linedash_yearend}-12"]
 linedash_month_timespan = ["01", "12"]
+linedash_year_options = [str(year) for year in range(int(linedash_year_timespan[0]), int(linedash_year_timespan[1])+1)]
 
 # bokeh server execution commands
 bat_execBokehApp = "START /MIN CMD.EXE /C exeBokehApp.bat"

dashboard/utilities/bokeh_line_dash.py

@@ -1,4 +1,5 @@
 import pickle
+import logging
 from beartype import beartype
 from typing import Union
 from bokeh.models import Select, CheckboxGroup, Div, Button, MultiSelect
@@ -8,6 +9,7 @@
 import cons
 from utilities.bokeh_line_data import bokeh_line_data
 from utilities.bokeh_line_plot import bokeh_line_plot
+from utilities.timeit import timeit
 
 @beartype
 def bokeh_line_dash():
@@ -21,21 +23,20 @@ def bokeh_line_dash():
     bokeh.layouts.row
         The interactive bokeh line dashboard
     """
+    logging.info("Initialise line plot begin")
     with open(cons.preaggregate_data_fpath, "rb") as handle:
         pre_agg_data_dict = pickle.load(handle)
     # generate bokeh data for line plot
-    bokeh_line_data_dict = bokeh_line_data(pre_agg_data_dict)
+    bokeh_line_data_params = {"pre_agg_data_dict":pre_agg_data_dict}
+    bokeh_line_data_dict = timeit(func=bokeh_line_data, params=bokeh_line_data_params)
     # create bokeh plot
-    line_plot = bokeh_line_plot(
-        bokeh_line_data_dict,
-        col=cons.col_default,
-        stat=cons.stat_default,
-        agg_level=cons.line_agg_level_default,
-        selection=cons.counties,
-    )
+    bokeh_line_plot_params = {"bokeh_data_dict":bokeh_line_data_dict, "col":cons.col_default, "stat":cons.stat_default, "agg_level":cons.line_agg_level_default, "selection":cons.counties}
+    line_plot = timeit(func=bokeh_line_plot, params=bokeh_line_plot_params)
+    logging.info("Initialise line plot end")
 
     # create call back function for bokeh dashboard interaction
     def callback_line_plot(attr, old, new):
+        logging.info("Callback line plot begin")
         # extract new selector value
         agg_level = line_agg_level_selector.value
         col = line_col_selector.value
@@ -44,15 +45,11 @@ def callback_line_plot(attr, old, new):
         for i in line_county_multiselect.value:
             selection.append(cons.counties[int(i)])
         # update bokeh plot
-        line_plot = bokeh_line_plot(
-            bokeh_line_data_dict,
-            col=col,
-            stat=stat,
-            agg_level=agg_level,
-            selection=selection,
-        )
+        bokeh_line_plot_params = {"bokeh_data_dict":bokeh_line_data_dict, "col":col, "stat":stat, "agg_level":agg_level, "selection":selection}
+        line_plot = timeit(func=bokeh_line_plot, params=bokeh_line_plot_params)
         # reassign bokeh plot to bokeh dashboard
         dashboard_line.children[1] = line_plot
+        logging.info("Callback line plot end")
 
     def callback_multiselect_selectall():
         line_county_multiselect.value = cons.counties_values

dashboard/utilities/bokeh_map_dash.py

@@ -1,11 +1,13 @@
 import pickle
+import logging
 from bokeh.models import Select, Div, CheckboxButtonGroup
 from bokeh.layouts import column, row
 from beartype import beartype
 
 import cons
 from utilities.bokeh_map_data import bokeh_map_data
 from utilities.bokeh_map_plot import bokeh_map_plot
+from utilities.timeit import timeit
 
 @beartype
 def bokeh_map_dash():
@@ -19,37 +21,36 @@ def bokeh_map_dash():
     bokeh.layouts.row
         The interactive bokeh map dashboard
     """
+    logging.info("Initialise map plot begin")
     with open(cons.map_data_fpath, "rb") as handle:
-        map_data_dict = pickle.load(handle)
+        map_data = pickle.load(handle)
     with open(cons.points_data_fpath, "rb") as handle:
         station_data = pickle.load(handle)
     # generate bokeh data for map plot
-    bokeh_map_data_dict, pointgeosource = bokeh_map_data(map_data_dict, station_data)
+    bokeh_map_data_params = {"map_data":map_data,"station_data":station_data,"col":cons.col_default,"stat":cons.stat_default,"year":cons.linedash_year_timespan[1]}
+    bokeh_map_data_dict = timeit(func=bokeh_map_data, params=bokeh_map_data_params)
     # create bokeh map plot
-    map_plot = bokeh_map_plot(
-        bokeh_map_data_dict=bokeh_map_data_dict,
-        pointgeosource=pointgeosource,
-        col=cons.col_default,
-        stat=cons.stat_default,
-        show_stations=cons.show_stations_default,
-    )
+    bokeh_map_plot_params = {"bokeh_map_data_dict":bokeh_map_data_dict,"show_stations":cons.show_stations_default}
+    map_plot = timeit(func=bokeh_map_plot, params=bokeh_map_plot_params)
+    logging.info("Initialise map plot end")
 
     # create call back function for bokeh dashboard interaction
     def callback_map_plot(attr, old, new):
+        logging.info("Callback map plot begin")
         # extract new selector value
         col = map_col_selector.value
         stat = map_stat_selector.value
+        year = map_year_selector.value
         show_stations = toggle_stations.active
+        # update bokeh data
+        bokeh_map_data_params = {"map_data":map_data,"station_data":station_data,"col":col,"stat":stat,"year":year}
+        bokeh_map_data_dict = timeit(func=bokeh_map_data, params=bokeh_map_data_params)
         # update bokeh plot
-        map_plot = bokeh_map_plot(
-            bokeh_map_data_dict=bokeh_map_data_dict,
-            pointgeosource=pointgeosource,
-            col=col,
-            stat=stat,
-            show_stations=show_stations,
-        )
+        bokeh_map_plot_params = {"bokeh_map_data_dict":bokeh_map_data_dict,"show_stations":show_stations}
+        map_plot = timeit(func=bokeh_map_plot, params=bokeh_map_plot_params)
         # reassign bokeh plot to bokeh dashboard
         dashboard_map.children[1] = map_plot
+        logging.info("Callback map plot end")
 
     # set up selectors for bokeh map plot
     map_col_selector = Select(
@@ -68,6 +69,14 @@ def callback_map_plot(attr, old, new):
         height=60,
         aspect_ratio=10,
     )
+    map_year_selector = Select(
+        title="Year:",
+        value=cons.linedash_year_timespan[1],
+        options=cons.linedash_year_options,
+        width=120,
+       height=60,
+        aspect_ratio=10,
+    )
     toggle_stations = CheckboxButtonGroup(
         labels=["Toggle Stations"], 
         active=[], 
@@ -77,11 +86,10 @@ def callback_map_plot(attr, old, new):
     map_col_selector.on_change("value", callback_map_plot)
     map_stat_selector.on_change("value", callback_map_plot)
     toggle_stations.on_change("active", callback_map_plot)
+    map_year_selector.on_change("value", callback_map_plot)
 
     # structure dashboard map plot
-    widgets_map = column(
-        toggle_stations, map_col_selector, map_stat_selector
-    )
+    widgets_map = column(toggle_stations, map_col_selector, map_stat_selector, map_year_selector)
     dashboard_map = row(widgets_map, map_plot)
 
     return dashboard_map

dashboard/utilities/bokeh_map_data.py

@@ -7,42 +7,58 @@
 
 @beartype
 def bokeh_map_data(
-    map_data_dict:dict, 
-    station_data:gpd.GeoDataFrame
+    map_data:gpd.GeoDataFrame, 
+    station_data:gpd.GeoDataFrame,
+    col:str,
+    stat:str,
+    year:str
     ):
     """Generates the data used in the bokeh map plot.
 
     Parameters
     ----------
-    map_data_dict : dict
+    map_data : gpd.GeoDataFrame
         The aggregated geopandas data to be transformed into aggregated bokeh data objects for visualisation
     station_data : geopandas.DataFrame
         The geospatial Met Eireann station data to be transformed into a bokeh data object for visualisation
+    col : str
+        The climate measure category to plot in the interactive bokeh heatmap
+    stat : str
+        The aggregated statistic to plot in the interactive bokeh heatmap
+    year : str
+        The year of the data to filter for
 
     Returns
     -------
-    dict, bokeh.models.GeoJSONDataSource
+    dict
         The aggregated and station bokeh data objects to visualise
     """
     bokeh_map_data_dict = {}
-    for stat, map_data in map_data_dict.items():
-        tmp_data_dict = {}
-        # split data into missing and nonmissing
-        nonmiss_map_data = map_data[map_data[cons.col_options].notnull().any(axis=1)]
-        miss_map_data = map_data[~map_data[cons.col_options].notnull().any(axis=1)]
-        # Input GeoJSON source that contains features for plotting
-        missgeosource = GeoJSONDataSource(geojson=miss_map_data.to_json())
-        nonmissgeosource = GeoJSONDataSource(geojson=nonmiss_map_data.to_json())
-        # assign data to temp dict
-        tmp_data_dict["map_data"] = map_data
-        tmp_data_dict["nonmiss_map_data"] = nonmiss_map_data
-        tmp_data_dict["miss_map_data"] = miss_map_data
-        tmp_data_dict["missgeosource"] = missgeosource
-        tmp_data_dict["nonmissgeosource"] = nonmissgeosource
-        # assign temp data dict to bokeh data dict
-        bokeh_map_data_dict[stat] = tmp_data_dict
+    # filter for year
+    row_filter = (map_data['year'].astype(str) == year) & (map_data['stat'].astype(str) == stat)
+    sub_cols = ['county', 'geometry', 'year', 'stat', col]
+    map_data_sub = map_data.loc[row_filter, sub_cols]
+    # split data into missing and nonmissing
+    nonmiss_map_data = map_data_sub[map_data_sub.notnull().all(axis=1)]
+    miss_map_data = map_data_sub[map_data_sub.isnull().any(axis=1)]
+    # Input GeoJSON source that contains features for plotting
+    missgeosource = GeoJSONDataSource(geojson=miss_map_data.to_json())
+    nonmissgeosource = GeoJSONDataSource(geojson=nonmiss_map_data.to_json())
     pointgeosource = GeoJSONDataSource(geojson=station_data.to_json())
+    # calculate colour bar range
+    color_mapper_low = nonmiss_map_data[col].min()
+    color_mapper_high = nonmiss_map_data[col].max()
+    # assign data to temp dict
+    bokeh_map_data_dict["col"] = col
+    bokeh_map_data_dict["map_data"] = map_data
+    bokeh_map_data_dict["nonmiss_map_data"] = nonmiss_map_data
+    bokeh_map_data_dict["miss_map_data"] = miss_map_data
+    bokeh_map_data_dict["missgeosource"] = missgeosource
+    bokeh_map_data_dict["nonmissgeosource"] = nonmissgeosource
+    bokeh_map_data_dict["pointgeosource"] = pointgeosource
+    bokeh_map_data_dict["color_mapper_low"] = color_mapper_low
+    bokeh_map_data_dict["color_mapper_high"] = color_mapper_high
     # pickle the bokeh map data dictionary to disk
     # with open(cons.bokeh_map_data_fpath, 'wb') as f:
     #    pickle.dump(bokeh_map_data_dict, f, protocol = pickle.HIGHEST_PROTOCOL)
-    return bokeh_map_data_dict, pointgeosource
+    return bokeh_map_data_dict

dashboard/utilities/bokeh_map_plot.py

@@ -8,9 +8,6 @@
 @beartype
 def bokeh_map_plot(
     bokeh_map_data_dict:dict, 
-    pointgeosource:GeoJSONDataSource, 
-    col:str, 
-    stat:str, 
     show_stations:list
     ) -> figure:
     """Generates the data used in the bokeh map plot.
@@ -19,12 +16,6 @@ def bokeh_map_plot(
     ----------
     bokeh_map_data_dict : dict
         A dictionary of bokeh aggregated data objects to construct the interactive bokeh heatmap with
-    pointgeosource : geopandas.DataFrame
-        The geospatial Met Eireann station data to overlay was red dots in the interactive bokeh heatmap
-    col : str
-        The climate measure category to plot in the interactive bokeh heatmap
-    stat : str
-        The aggregated statistic to plot in the interactive bokeh heatmap
     show_stations : list
         Whether to toggle the Met Eireann station data overlay in the interactive bokeh heatmap
 
@@ -37,13 +28,11 @@ def bokeh_map_plot(
     lightblue = Color("lightblue")
     steelblue = Color("steelblue")
     palette = tuple([col.get_hex() for col in lightblue.range_to(steelblue, 100)])
-    # palette = ("lightblue", "steelblue")
-    # instantiate LinearColorMapper that linearly maps numbers in a range, into a sequence of colors.
-    nonmiss_map_data = bokeh_map_data_dict[stat]["nonmiss_map_data"]
+    # instantiate LinearColorMapper that linearly maps numbers in a range, into a sequence of colours.
     color_mapper = LinearColorMapper(
         palette=palette,
-        low=nonmiss_map_data[col].min(),
-        high=nonmiss_map_data[col].max(),
+        low=bokeh_map_data_dict["color_mapper_low"],
+        high=bokeh_map_data_dict["color_mapper_high"],
     )
     # create color bar.
     color_bar = ColorBar(
@@ -57,9 +46,7 @@ def bokeh_map_plot(
         major_label_text_font_size="18px",
     )
     # create underlying figure object
-    map_plot = figure(
-        toolbar_location="below", output_backend="webgl", **cons.FIG_SETTING
-    )
+    map_plot = figure(toolbar_location="below", output_backend="webgl", **cons.FIG_SETTING)
     map_plot.axis.visible = False
     map_plot.xgrid.grid_line_color = None
     map_plot.ygrid.grid_line_color = None
@@ -70,9 +57,9 @@ def bokeh_map_plot(
     map_plot.title.text_font_style = "bold"
     map_plot.title.text_font_size = "22px"
     # add patches to render states with no aggregate data
-    missgeosource = bokeh_map_data_dict[stat]["missgeosource"]
+    missgeosource = bokeh_map_data_dict['missgeosource']#[year]['miss_map_dataview']
     misscounties = map_plot.patches(
-        "xs", "ys", source=missgeosource, fill_color="white", **cons.MAP_SETTINGS
+        xs="xs", ys="ys", source=missgeosource, fill_color="slategray", **cons.MAP_SETTINGS
     )
     map_plot.add_tools(
         HoverTool(
@@ -83,26 +70,26 @@ def bokeh_map_plot(
         )
     )
     # add patches to render states with aggregate data
-    nonmissgeosource = bokeh_map_data_dict[stat]["nonmissgeosource"]
+    nonmissgeosource = bokeh_map_data_dict['nonmissgeosource']
     nonmisscounties = map_plot.patches(
-        "xs",
-        "ys",
+        xs="xs",
+        ys="ys",
         source=nonmissgeosource,
-        fill_color={"field": col, "transform": color_mapper},
+        fill_color={"field": bokeh_map_data_dict['col'], "transform": color_mapper},
         **cons.MAP_SETTINGS,
     )
     map_plot.add_tools(
         HoverTool(
             renderers=[nonmisscounties],
-            tooltips=[("County Name", "@county"), ("County Value", f"@{col}")],
+            tooltips=[("County Name", "@county"), ("County Value", f"@{bokeh_map_data_dict['col']}")],
             attachment="left",
             mode="mouse",
         )
     )
     # add points to render stations
     if show_stations == [0]:
         stationpoints = map_plot.scatter(
-            "x", "y", source=pointgeosource, color="red", size=8, alpha=0.3
+            x="x", y="y", source=bokeh_map_data_dict['pointgeosource'], color="red", size=8, alpha=0.3
         )
         map_plot.add_tools(
             HoverTool(

dashboard/arch/timeit.py → dashboard/utilities/timeit.py

@@ -1,5 +1,6 @@
 import time
 import numpy as np
+import logging
 
 
 def timeit(func, params, itr=1, digits=3):
@@ -27,5 +28,8 @@ def timeit(func, params, itr=1, digits=3):
         t1 = time.time()
         tres = t1 - t0
     eres = round(np.mean(tres), digits)
-    print(f"Mean Execution Time: {eres} seconds")
+    logging_message=f"Execution Time for {str(func)}: {eres} seconds"
+    if itr>1:
+        logging_message=f"Mean {logging_message}"
+    logging.info(logging_message)
     return res