Merge pull request #138 from agoenergy/unified_costs_calculation_per_…

…dimension

Update cost results presentation

app/layout_elements.py

@@ -0,0 +1,60 @@
+# -*- coding: utf-8 -*-
+"""Layout elements that get reused in several tabs."""
+import pandas as pd
+import streamlit as st
+
+from app.plot_functions import create_bar_chart_costs
+from app.ptxboa_functions import config_number_columns
+
+
+def display_costs(
+    df_costs: pd.DataFrame, key: str, titlestring: str, key_suffix: str = ""
+):
+    """Display costs as table and bar chart."""
+    key_suffix = key_suffix.lower().replace(" ", "_")
+    st.subheader(titlestring)
+    c1, c2 = st.columns([1, 5])
+    with c1:
+        # filter data:
+        df_res = df_costs.copy()
+
+        # select filter:
+        show_which_data = st.radio(
+            "Select elements to display:",
+            ["All", "Manual select"],
+            index=0,
+            key=f"show_which_data_{key}_{key_suffix}",
+        )
+
+        # apply filter:
+        if show_which_data == "Manual select":
+            ind_select = st.multiselect(
+                "Select regions:",
+                df_res.index.values,
+                default=df_res.index.values,
+                key=f"select_data_{key}_{key_suffix}",
+            )
+            df_res = df_res.loc[ind_select]
+
+        # sort:
+        sort_ascending = st.toggle(
+            "Sort by total costs?",
+            value=True,
+            key=f"sort_data_{key}_{key_suffix}",
+        )
+        if sort_ascending:
+            df_res = df_res.sort_values(["Total"], ascending=True)
+    with c2:
+        # create graph:
+        fig = create_bar_chart_costs(
+            df_res,
+            current_selection=st.session_state[key],
+        )
+        st.plotly_chart(fig, use_container_width=True)
+
+    with st.expander("**Data**"):
+        column_config = config_number_columns(
+            df_res, format=f"%.1f {st.session_state['output_unit']}"
+        )
+        st.dataframe(df_res, use_container_width=True, column_config=column_config)
+    return None

app/plot_functions.py

@@ -9,7 +9,11 @@
 import plotly.graph_objects as go
 import streamlit as st
 
-from app.ptxboa_functions import remove_subregions, subset_and_pivot_input_data
+from app.ptxboa_functions import (
+    remove_subregions,
+    select_subregions,
+    subset_and_pivot_input_data,
+)
 from ptxboa.api import PtxboaAPI
 
 
@@ -226,7 +230,7 @@ def _choropleth_map_deep_dive_country(
     if custom_data_func_kwargs is None:
         custom_data_func_kwargs = {}
     # subsetting 'df' for the selected deep dive country
-    df = df.copy().loc[df.index.str.startswith(f"{deep_dive_country} ("), :]
+    df = select_subregions(df, deep_dive_country)
     # need to calculate custom data befor is03166 column is appended.
     hover_data = custom_data_func(df, **custom_data_func_kwargs)
     # get dataframe with info about iso 3166-2 codes and map them to res_costs

app/ptxboa_functions.py

@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 """Utility functions for streamlit app."""
+from typing import Literal
 
 import pandas as pd
 import streamlit as st
@@ -83,19 +84,28 @@ def calculate_results_list(
     for parameter in parameter_list:
         settings2 = settings.copy()
         settings2[parameter_to_change] = parameter
-        res_single = calculate_results_single(api, settings2, user_data=user_data)
+        res_single = calculate_results_single(
+            api,
+            settings2,
+            user_data=st.session_state["user_changes_df"],
+        )
         res_list.append(res_single)
     res_details = pd.concat(res_list)
 
-    return aggregate_costs(res_details)
+    return aggregate_costs(res_details, parameter_to_change)
 
 
-def aggregate_costs(res_details: pd.DataFrame) -> pd.DataFrame:
+def aggregate_costs(
+    res_details: pd.DataFrame, parameter_to_change: str
+) -> pd.DataFrame:
     """Aggregate detailed costs."""
     # Exclude levelized costs:
     res = res_details.loc[res_details["cost_type"] != "LC"]
     res = res.pivot_table(
-        index="region", columns="process_type", values="values", aggfunc="sum"
+        index=parameter_to_change,
+        columns="process_type",
+        values="values",
+        aggfunc="sum",
     )
     # calculate total costs:
     res["Total"] = res.sum(axis=1)
@@ -185,6 +195,27 @@ def remove_subregions(api: PtxboaAPI, df: pd.DataFrame, country_name: str):
     return df
 
 
+def select_subregions(
+    df: pd.DataFrame, deep_dive_country: Literal["Argentina", "Morocco", "South Africa"]
+) -> pd.DataFrame:
+    """
+    Only select rows corresponding to subregions of a deep dive country.
+
+    Parameters
+    ----------
+    df : pd.DataFrame
+        pandas DataFrame with list of regions as index.
+    deep_dive_country : str in {"Argentina", "Morocco", "South Africa"}
+
+
+    Returns
+    -------
+    pd.DataFrame
+    """
+    df = df.copy().loc[df.index.str.startswith(f"{deep_dive_country} ("), :]
+    return df
+
+
 def reset_user_changes():
     """Reset all user changes."""
     if (

app/tab_dashboard.py

@@ -1,13 +1,17 @@
 # -*- coding: utf-8 -*-
 """Content of dashboard tab."""
+import pandas as pd
 import streamlit as st
 from plotly.subplots import make_subplots
 
+from app.layout_elements import display_costs
 from app.plot_functions import (
     create_bar_chart_costs,
     create_box_plot,
     plot_costs_on_map,
 )
+from app.ptxboa_functions import remove_subregions
+from ptxboa.api import PtxboaAPI
 
 
 def _create_infobox(context_data: dict):
@@ -30,7 +34,14 @@ def write_info(info):
     write_info(info4)
 
 
-def content_dashboard(api, res_costs: dict, context_data: dict):
+def content_dashboard(
+    api: PtxboaAPI,
+    costs_per_region: pd.DataFrame,
+    costs_per_scenario: pd.DataFrame,
+    costs_per_res_gen: pd.DataFrame,
+    costs_per_chain: pd.DataFrame,
+    context_data: dict,
+):
     with st.expander("What is this?"):
         st.markdown(
             """
@@ -39,6 +50,8 @@ def content_dashboard(api, res_costs: dict, context_data: dict):
 regional distribution of total costs across supply regions
 - a split-up of costs by category for your chosen supply region
 - key information on your chosen demand country.
+- total cost and cost components for different supply countries, scenarios,
+renewable electricity sources and process chains.
 
 Switch to other tabs to explore data and results in more detail!
             """
@@ -48,14 +61,16 @@ def content_dashboard(api, res_costs: dict, context_data: dict):
 
     with c_1:
         fig_map = plot_costs_on_map(
-            api, res_costs, scope="world", cost_component="Total"
+            api, costs_per_region, scope="world", cost_component="Total"
         )
         st.plotly_chart(fig_map, use_container_width=True)
 
     with c_2:
         # create box plot and bar plot:
-        fig1 = create_box_plot(res_costs)
-        filtered_data = res_costs[res_costs.index == st.session_state["region"]]
+        fig1 = create_box_plot(costs_per_region)
+        filtered_data = costs_per_region[
+            costs_per_region.index == st.session_state["region"]
+        ]
         fig2 = create_bar_chart_costs(filtered_data)
         doublefig = make_subplots(rows=1, cols=2, shared_yaxes=True)
 
@@ -70,3 +85,17 @@ def content_dashboard(api, res_costs: dict, context_data: dict):
         st.plotly_chart(doublefig, use_container_width=True)
 
         _create_infobox(context_data)
+
+    display_costs(
+        remove_subregions(api, costs_per_region, st.session_state["country"]),
+        "region",
+        "Costs by region:",
+    )
+
+    display_costs(costs_per_scenario, "scenario", "Costs by data scenario:")
+
+    display_costs(
+        costs_per_res_gen, "res_gen", "Costs by renewable electricity source:"
+    )
+
+    display_costs(costs_per_chain, "chain", "Costs by supply chain:")

app/tab_deep_dive_countries.py

@@ -4,19 +4,20 @@
 import plotly.express as px
 import streamlit as st
 
+from app.layout_elements import display_costs
 from app.plot_functions import plot_costs_on_map, plot_input_data_on_map
-from app.ptxboa_functions import display_and_edit_data_table
+from app.ptxboa_functions import display_and_edit_data_table, select_subregions
 from ptxboa.api import PtxboaAPI
 
 
-def content_deep_dive_countries(api: PtxboaAPI, res_costs: pd.DataFrame) -> None:
+def content_deep_dive_countries(api: PtxboaAPI, costs_per_region: pd.DataFrame) -> None:
     """Create content for the "costs by region" sheet.
 
     Parameters
     ----------
     api : :class:`~ptxboa.api.PtxboaAPI`
         an instance of the api class
-    res_costs : pd.DataFrame
+    costs_per_region : pd.DataFrame
         Results.
 
     Output
@@ -41,9 +42,18 @@ def content_deep_dive_countries(api: PtxboaAPI, res_costs: pd.DataFrame) -> None
         "Select country:", ["Argentina", "Morocco", "South Africa"], horizontal=True
     )
 
-    fig_map = plot_costs_on_map(api, res_costs, scope=ddc, cost_component="Total")
+    fig_map = plot_costs_on_map(
+        api, costs_per_region, scope=ddc, cost_component="Total"
+    )
     st.plotly_chart(fig_map, use_container_width=True)
 
+    display_costs(
+        select_subregions(costs_per_region, ddc),
+        key="region",
+        titlestring="Costs per subregion",
+        key_suffix=ddc,
+    )
+
     st.subheader("Full load hours of renewable generation")
     input_data = api.get_input_data(st.session_state["scenario"])
     # filter data: