huff 1.1.2__py3-none-any.whl → 1.2.0__py3-none-any.whl

huff/models.py

@@ -4,8 +4,8 @@
 # Author:      Thomas Wieland 
 #              ORCID: 0000-0001-5168-9846
 #              mail: geowieland@googlemail.com              
-# Version:     1.1.2
-# Last update: 2025-05-03 13:29
+# Version:     1.2.0
+# Last update: 2025-05-14 18:33
 # Copyright (c) 2025 Thomas Wieland
 #-----------------------------------------------------------------------
 
@@ -14,6 +14,9 @@ import pandas as pd
 import geopandas as gp
 import numpy as np
 import time
+from statsmodels.formula.api import ols
+from shapely.geometry import Point
+from shapely import wkt
 from huff.ors import Client, TimeDistanceMatrix, Isochrone
 from huff.gistools import overlay_difference, distance_matrix
 
@@ -24,12 +27,14 @@ class CustomerOrigins:
         self,
         geodata_gpd, 
         geodata_gpd_original, 
-        metadata
+        metadata,
+        isochrones_gdf
         ):
 
         self.geodata_gpd = geodata_gpd
         self.geodata_gpd_original = geodata_gpd_original
         self.metadata = metadata
+        self.isochrones_gdf = isochrones_gdf
 
     def get_geodata_gpd(self):
 
@@ -42,6 +47,10 @@ class CustomerOrigins:
     def get_metadata(self):
         
         return self.metadata
+    
+    def get_isochrones(self):
+
+        return self.isochrones_gdf
 
     def summary(self):
 
@@ -63,6 +72,11 @@ class CustomerOrigins:
         print("Unique ID column           " + metadata["unique_id"])
         print("Input CRS                  " + str(metadata["crs_input"]))
 
+        if self.isochrones_gdf is None:
+            print("Including isochrones       NO")
+        else:
+            print("Including isochrones       YES")
+
         return metadata
     
     def define_marketsize(
@@ -97,27 +111,77 @@ class CustomerOrigins:
 
         return self
 
+    def isochrones(
+        self,
+        segments_minutes: list = [5, 10, 15],
+        range_type: str = "time",
+        intersections: str = "true",
+        profile: str = "driving-car",
+        donut: bool = True,
+        ors_server: str = "https://api.openrouteservice.org/v2/",
+        ors_auth: str = None,
+        timeout: int = 10,
+        delay: int = 1,
+        save_output: bool = True,
+        output_filepath: str = "customer_origins_isochrones.shp",
+        output_crs: str = "EPSG:4326"
+        ):
+
+        geodata_gpd = self.get_geodata_gpd()
+        metadata = self.get_metadata()
+
+        isochrones_gdf = get_isochrones( 
+            geodata_gpd = geodata_gpd,
+            unique_id_col = metadata["unique_id"],
+            segments_minutes = segments_minutes,
+            range_type = range_type,
+            intersections = intersections,
+            profile = profile,
+            donut = donut,
+            ors_server = ors_server,
+            ors_auth = ors_auth,
+            timeout = timeout,
+            delay = delay,
+            save_output = save_output,
+            output_filepath = output_filepath,
+            output_crs = output_crs
+            )
+
+        self.isochrones_gdf = isochrones_gdf
+
+        return self
+
+
 class SupplyLocations:
 
     def __init__(
         self,
         geodata_gpd, 
         geodata_gpd_original, 
-        metadata
+        metadata,
+        isochrones_gdf
         ):
 
         self.geodata_gpd = geodata_gpd
         self.geodata_gpd_original = geodata_gpd_original
         self.metadata = metadata
+        self.isochrones_gdf = isochrones_gdf
 
     def get_geodata_gpd(self):
+
         return self.geodata_gpd
 
     def get_geodata_gpd_original(self):
+
         return self.geodata_gpd_original
 
     def get_metadata(self):
+
         return self.metadata
+    
+    def get_isochrones_gdf(self):
+
+        return self.isochrones_gdf
 
     def summary(self):
 
@@ -139,6 +203,11 @@ class SupplyLocations:
         print("Unique ID column      " + metadata["unique_id"])
         print("Input CRS             " + str(metadata["crs_input"]))
 
+        if self.isochrones_gdf is None:
+            print("Including isochrones  NO")
+        else:
+            print("Including isochrones  YES")
+
         return metadata
 
     def define_attraction(
@@ -239,86 +308,44 @@ class SupplyLocations:
     
     def isochrones(
         self,
-        segments: list = [900, 600, 300],
+        segments_minutes: list = [5, 10, 15],
         range_type: str = "time",
         intersections: str = "true",
         profile: str = "driving-car",
         donut: bool = True,
         ors_server: str = "https://api.openrouteservice.org/v2/",
         ors_auth: str = None,
-        timeout = 10,
-        delay = 1,
+        timeout: int = 10,
+        delay: int = 1,
         save_output: bool = True,
-        output_filepath: str = "isochrones.shp",
+        output_filepath: str = "supply_locations_isochrones.shp",
         output_crs: str = "EPSG:4326"
         ):
 
         geodata_gpd = self.get_geodata_gpd()
         metadata = self.get_metadata()
 
-        coords = [(point.x, point.y) for point in geodata_gpd.geometry]
-        
-        unique_id_col = metadata["unique_id"]
-        unique_id_values = geodata_gpd[unique_id_col].values
-
-        ors_client = Client(
-            server = ors_server,
-            auth = ors_auth
+        isochrones_gdf = get_isochrones( 
+            geodata_gpd = geodata_gpd,
+            unique_id_col = metadata["unique_id"],
+            segments_minutes = segments_minutes,
+            range_type = range_type,
+            intersections = intersections,
+            profile = profile,
+            donut = donut,
+            ors_server = ors_server,
+            ors_auth = ors_auth,
+            timeout = timeout,
+            delay = delay,
+            save_output = save_output,
+            output_filepath = output_filepath,
+            output_crs = output_crs
             )
-        
-        isochrones_gdf = gp.GeoDataFrame(columns=[unique_id_col, "geometry"])
-        
-        i = 0
-
-        for x, y in coords:
-            
-            isochrone_output = ors_client.isochrone(
-                locations = [[x, y]],
-                segments = segments,
-                range_type = range_type,
-                intersections = intersections,
-                profile = profile,
-                timeout = timeout,
-                save_output = False,
-                output_crs = output_crs
-                )
-            
-            if isochrone_output.status_code != 200:
-                continue        
-            
-            isochrone_gdf = isochrone_output.get_isochrones_gdf()
-            
-            if donut:
-                isochrone_gdf = overlay_difference(
-                    polygon_gdf = isochrone_gdf, 
-                    sort_col = "segment"
-                    )
-                
-            time.sleep(delay)
-
-            isochrone_gdf[unique_id_col] = unique_id_values[i]
-
-            isochrones_gdf = pd.concat(
-                [
-                    isochrones_gdf, 
-                    isochrone_gdf
-                    ], 
-                ignore_index=True
-                )
-            
-            i = i+1
 
-        isochrones_gdf.set_crs(
-            output_crs, 
-            allow_override=True, 
-            inplace=True
-            )
-            
-        if save_output:
+        self.isochrones_gdf = isochrones_gdf
 
-            isochrones_gdf.to_file(filename = output_filepath)
+        return self
 
-        return isochrones_gdf
 
 class InteractionMatrix:
 
@@ -360,15 +387,16 @@ class InteractionMatrix:
         else:
             print("Market size column " + customer_origins_metadata["marketsize_col"])
         print("----------------------------------")
-        print("Weights")
+        print("Partial utilities")
+        print("                   Weights")
         if supply_locations_metadata["weighting"][0]["func"] is None and supply_locations_metadata["weighting"][0]["param"] is None:
-            print("Gamma              not defined")
+            print("Attraction         not defined")
         else:
-            print("Gamma              " + str(supply_locations_metadata["weighting"][0]["param"]) + " (" + supply_locations_metadata["weighting"][0]["func"] + ")")
+            print("Attraction         " + str(supply_locations_metadata["weighting"][0]["param"]) + " (" + supply_locations_metadata["weighting"][0]["func"] + ")")
         if customer_origins_metadata["weighting"][0]["func"] is None and customer_origins_metadata["weighting"][0]["param"] is None:
-            print("Lambda             not defined")
+            print("Transport costs    not defined")
         else:
-            print("Lambda            " + str(customer_origins_metadata["weighting"][0]["param"]) + " (" + customer_origins_metadata["weighting"][0]["func"] + ")")     
+            print("Transport costs    " + str(customer_origins_metadata["weighting"][0]["param"]) + " (" + customer_origins_metadata["weighting"][0]["func"] + ")")     
         print("----------------------------------")
 
     def transport_costs(
@@ -506,11 +534,11 @@ class InteractionMatrix:
         if attraction_weighting["func"] == "power":
             interaction_matrix_df["A_j_weighted"] = interaction_matrix_df["A_j"] ** attraction_weighting["param"]
         elif tc_weighting["func"] == "exponential":
-            interaction_matrix_df["A_j_weighted"] = np.exp(attraction_weighting["param"] * interaction_matrix_df['A_j'])
+            interaction_matrix_df["A_j_weighted"] = np.exp(attraction_weighting["param"] * interaction_matrix_df["A_j"])
         else:
             raise ValueError ("Attraction weighting is not defined.")
         
-        interaction_matrix_df["U_ij"] = interaction_matrix_df["A_j_weighted"]/interaction_matrix_df["t_ij_weighted"]
+        interaction_matrix_df["U_ij"] = interaction_matrix_df["A_j_weighted"]*interaction_matrix_df["t_ij_weighted"]
         
         interaction_matrix_df = interaction_matrix_df.drop(columns=['A_j_weighted', 't_ij_weighted'])
 
@@ -591,13 +619,11 @@ class InteractionMatrix:
         cols: list = ["A_j", "t_ij"]
         ):
 
-        """ MCI model log-centering transformation """
-        
         cols = cols + ["p_ij"]
 
         interaction_matrix_df = self.interaction_matrix_df
 
-        interaction_matrix_df = mci_transformation(
+        interaction_matrix_df = log_centering_transformation(
             df = interaction_matrix_df,
             ref_col = "i",
             cols = cols
@@ -607,6 +633,87 @@ class InteractionMatrix:
 
         return self
 
+    def mci_fit(
+        self,
+        cols: list = ["A_j", "t_ij"],
+        alpha = 0.05
+        ):
+
+        supply_locations = self.get_supply_locations()
+        supply_locations_metadata = supply_locations.get_metadata()
+
+        customer_origins = self.get_customer_origins()
+        customer_origins_metadata = customer_origins.get_metadata()
+
+        interaction_matrix_df = self.get_interaction_matrix_df()
+
+        cols_t = [col + "__LCT" for col in cols]
+
+        if "p_ij__LCT" not in interaction_matrix_df.columns:
+            interaction_matrix = self.mci_transformation(
+                cols = cols
+                )
+            interaction_matrix_df = self.get_interaction_matrix_df()
+
+        mci_formula = f'p_ij__LCT ~ {" + ".join(cols_t)} -1'
+
+        mci_ols_model = ols(mci_formula, data = interaction_matrix_df).fit()
+
+        mci_ols_coefficients = mci_ols_model.params
+        mci_ols_coef_standarderrors = mci_ols_model.bse
+        mci_ols_coef_t = mci_ols_model.tvalues
+        mci_ols_coef_p = mci_ols_model.pvalues
+        mci_ols_coef_ci = mci_ols_model.conf_int(alpha = alpha)
+
+        coefs = {}
+        for i, col in enumerate(cols_t):
+            coefs[i] = {
+                "Coefficient": col[:-5],
+                "Estimate": float(mci_ols_coefficients[col]),
+                "SE": float(mci_ols_coef_standarderrors[col]),
+                "t": float(mci_ols_coef_t[col]),
+                "p": float(mci_ols_coef_p[col]),
+                "CI_lower": float(mci_ols_coef_ci.loc[col, 0]),
+                "CI_upper": float(mci_ols_coef_ci.loc[col, 1]),
+                }
+
+        customer_origins_metadata["weighting"][0] = {
+            "func": "power",
+            "param": mci_ols_coefficients["t_ij__LCT"]
+            }
+
+        coefs2 = coefs.copy()
+        for key, value in list(coefs2.items()):
+            if value["Coefficient"] == "t_ij":
+                del coefs2[key]
+
+        for key, value in coefs2.items():
+            supply_locations_metadata["weighting"][key] = {
+                "func": "power",
+                "param": value["Estimate"]
+            }
+
+            supply_locations_metadata["attraction_col"].append(None)
+            supply_locations_metadata["attraction_col"][key] = value["Coefficient"]
+
+        customer_origins.metadata = customer_origins_metadata
+        supply_locations.metadata = supply_locations_metadata
+        interaction_matrix = InteractionMatrix(
+            interaction_matrix_df,
+            customer_origins,
+            supply_locations
+            )
+        
+        mci_model = MCIModel(
+            interaction_matrix,
+            coefs,
+            mci_ols_model,
+            None
+            )
+        
+        return mci_model
+
+
 class HuffModel:
 
     def __init__(
@@ -662,19 +769,305 @@ class HuffModel:
         else:
             print("Market size column " + customer_origins_metadata["marketsize_col"])
         print("----------------------------------")
-        print("Weights")
+        print("Partial utilities")
+        print("                   Weights")
         if supply_locations_metadata["weighting"][0]["func"] is None and supply_locations_metadata["weighting"][0]["param"] is None:
-            print("Gamma              not defined")
+            print("Attraction         not defined")
         else:
-            print("Gamma              " + str(supply_locations_metadata["weighting"][0]["param"]) + " (" + supply_locations_metadata["weighting"][0]["func"] + ")")
+            print("Attraction         " + str(supply_locations_metadata["weighting"][0]["param"]) + " (" + supply_locations_metadata["weighting"][0]["func"] + ")")
         if customer_origins_metadata["weighting"][0]["func"] is None and customer_origins_metadata["weighting"][0]["param"] is None:
-            print("Lambda             not defined")
+            print("Transport costs    not defined")
         else:
-            print("Lambda            " + str(customer_origins_metadata["weighting"][0]["param"]) + " (" + customer_origins_metadata["weighting"][0]["func"] + ")")     
+            print("Transport costs   " + str(customer_origins_metadata["weighting"][0]["param"]) + " (" + customer_origins_metadata["weighting"][0]["func"] + ")")     
         print("----------------------------------")
-          
+
+    def mci_fit(
+        self,
+        cols: list = ["A_j", "t_ij"],
+        alpha = 0.05
+        ):
+
+        interaction_matrix = self.interaction_matrix
+        
+        supply_locations = interaction_matrix.get_supply_locations()
+        supply_locations_metadata = supply_locations.get_metadata()
+
+        customer_origins = interaction_matrix.get_customer_origins()
+        customer_origins_metadata = customer_origins.get_metadata()
+
+        interaction_matrix_df = interaction_matrix.get_interaction_matrix_df()
+
+        cols_t = [col + "__LCT" for col in cols]
+
+        if "p_ij__LCT" not in interaction_matrix_df.columns:
+            interaction_matrix = interaction_matrix.mci_transformation(
+                cols = cols
+                )
+            interaction_matrix_df = interaction_matrix.get_interaction_matrix_df()
+
+        mci_formula = f'p_ij__LCT ~ {" + ".join(cols_t)} -1'
+
+        mci_ols_model = ols(mci_formula, data = interaction_matrix_df).fit()
+
+        mci_ols_coefficients = mci_ols_model.params
+        mci_ols_coef_standarderrors = mci_ols_model.bse
+        mci_ols_coef_t = mci_ols_model.tvalues
+        mci_ols_coef_p = mci_ols_model.pvalues
+        mci_ols_coef_ci = mci_ols_model.conf_int(alpha = alpha)
+
+        coefs = {}
+        for i, col in enumerate(cols_t):
+            coefs[i] = {
+                "Coefficient": col[:-5],
+                "Estimate": float(mci_ols_coefficients[col]),
+                "SE": float(mci_ols_coef_standarderrors[col]),
+                "t": float(mci_ols_coef_t[col]),
+                "p": float(mci_ols_coef_p[col]),
+                "CI_lower": float(mci_ols_coef_ci.loc[col, 0]),
+                "CI_upper": float(mci_ols_coef_ci.loc[col, 1]),
+                }
+
+        customer_origins_metadata["weighting"][0] = {
+            "func": "power",
+            "param": mci_ols_coefficients["t_ij__LCT"]
+            }
+
+        coefs2 = coefs.copy()
+        for key, value in list(coefs2.items()):
+            if value["Coefficient"] == "t_ij":
+                del coefs2[key]
+
+        for key, value in coefs2.items():
+            supply_locations_metadata["weighting"][(key)] = {
+                "func": "power",
+                "param": value["Estimate"]
+            }
+            supply_locations_metadata["attraction_col"][key] = value["Coefficient"]
+
+        customer_origins.metadata = customer_origins_metadata
+        supply_locations.metadata = supply_locations_metadata
+        interaction_matrix = InteractionMatrix(
+            interaction_matrix_df,
+            customer_origins,
+            supply_locations
+            )
+        
+        mci_model = MCIModel(
+            interaction_matrix,
+            coefs,
+            mci_ols_model,
+            None
+            )
+        
+        return mci_model
+
+
+class MCIModel:
+
+    def __init__(
+        self,
+        interaction_matrix: InteractionMatrix,
+        coefs: dict,
+        mci_ols_model,
+        market_areas_df
+        ):
+
+        self.interaction_matrix = interaction_matrix
+        self.coefs = coefs
+        self.mci_ols_model = mci_ols_model
+        self.market_areas_df = market_areas_df
+
+    def get_interaction_matrix_df(self):
+
+        interaction_matrix = self.interaction_matrix
+        interaction_matrix_df = interaction_matrix.get_interaction_matrix_df()
+
+        return interaction_matrix_df
+    
+    def get_supply_locations(self):
+
+        interaction_matrix = self.interaction_matrix
+        supply_locations = interaction_matrix.get_supply_locations()
+
+        return supply_locations
+
+    def get_customer_origins(self):
+
+        interaction_matrix = self.interaction_matrix
+        customer_origins = interaction_matrix.get_customer_origins()
+
+        return customer_origins
+    
+    def get_mci_ols_model(self):
+
+        return self.mci_ols_model
+    
+    def get_coefs_dict(self):
+
+        return self.coefs
+    
+    def get_market_areas_df(self):
+
+        return self.market_areas_df
+
+    def summary(self):
+
+        interaction_matrix = self.interaction_matrix
+        coefs = self.coefs
+
+        customer_origins_metadata = interaction_matrix.get_customer_origins().get_metadata()
+        supply_locations_metadata = interaction_matrix.get_supply_locations().get_metadata()
+
+        print("Multiplicative Competitive Interaction Model")
+        print("--------------------------------------------")
+        print("Supply locations   " + str(supply_locations_metadata["no_points"]))
+        print("Customer origins   " + str(customer_origins_metadata["no_points"]))
+        print("--------------------------------------------")
+        print("Partial utilities")
+
+        coefficients_rows = []
+        for key, value in coefs.items():
+            coefficient_name = value["Coefficient"]
+            if coefficient_name == "A_j":
+                coefficient_name = "Attraction"
+            if coefficient_name == "t_ij":
+                coefficient_name = "Transport costs"
+            coefficients_rows.append({
+                "": coefficient_name,
+                "Estimate": round(value["Estimate"], 3),
+                "SE": round(value["SE"], 3),
+                "t": round(value["t"], 3),
+                "p": round(value["p"], 3),
+                "CI lower": round(value["CI_lower"], 3),
+                "CI upper": round(value["CI_upper"], 3)
+            })
+        coefficients_df = pd.DataFrame(coefficients_rows)
+        
+        print (coefficients_df)
+
+        print("--------------------------------------------")
+
+    def utility(
+        self,
+        transformation = "LCT"
+        ):
+        
+        interaction_matrix = self.interaction_matrix
+        
+        interaction_matrix_df = interaction_matrix.get_interaction_matrix_df()
+
+        if interaction_matrix_df["t_ij"].isna().all():
+            raise ValueError ("Transport cost variable is not defined")
+        if interaction_matrix_df["A_j"].isna().all():
+            raise ValueError ("Attraction variable is not defined")
+
+        check_vars(
+            df = interaction_matrix_df,
+            cols = ["A_j", "t_ij"]
+            )
+
+        customer_origins_metadata = interaction_matrix.get_customer_origins().get_metadata()
+        
+        t_ij_weighting = customer_origins_metadata["weighting"][0]["param"]
+
+        if transformation == "ILCT":
+            mci_formula = f"{t_ij_weighting}*t_ij"
+        else:
+            mci_formula = f"t_ij**{t_ij_weighting}"
+        
+        supply_locations_metadata = interaction_matrix.get_supply_locations().get_metadata()
+        attraction_col = supply_locations_metadata["attraction_col"]
+        attraction_weighting = supply_locations_metadata["weighting"]
+
+        if transformation == "ILCT":
+            for key, value in attraction_weighting.items():
+                mci_formula = mci_formula + f" + {value['param']}*{attraction_col[key]}"
+        else:
+            for key, value in attraction_weighting.items():
+                mci_formula = mci_formula + f" * {attraction_col[key]}**{value['param']}"
+
+        interaction_matrix_df["U_ij"] = interaction_matrix_df.apply(lambda row: eval(mci_formula, {}, row.to_dict()), axis=1)
+
+        if transformation == "ILCT":
+            interaction_matrix_df["U_ij"] = np.exp(interaction_matrix_df["U_ij"])
+
+        self.interaction_matrix = interaction_matrix_df
+
+        return self
+    
+    def probabilities (self):
+
+        interaction_matrix_df = self.interaction_matrix_df
+
+        if interaction_matrix_df["U_ij"].isna().all():
+            self.utility()
+            interaction_matrix_df = self.interaction_matrix_df
+
+        utility_i = pd.DataFrame(interaction_matrix_df.groupby("i")["U_ij"].sum())
+        utility_i = utility_i.rename(columns = {"U_ij": "U_i"})
+
+        interaction_matrix_df = interaction_matrix_df.merge(
+            utility_i,
+            left_on="i",
+            right_on="i",
+            how="inner"
+            )
+
+        interaction_matrix_df["p_ij"] = (interaction_matrix_df["U_ij"]) / (interaction_matrix_df["U_i"])
+
+        interaction_matrix_df = interaction_matrix_df.drop(columns=["U_i"])
+
+        self.interaction_matrix_df = interaction_matrix_df
+
+        return self
+        
+    def flows (self):
+
+        interaction_matrix_df = self.interaction_matrix_df
+
+        if interaction_matrix_df["C_i"].isna().all():
+            raise ValueError ("Market size column in customer origins not defined. Use CustomerOrigins.define_marketsize()")
+
+        check_vars(
+            df = interaction_matrix_df,
+            cols = ["C_i"]
+            )
+
+        if interaction_matrix_df["p_ij"].isna().all():
+            self.probabilities()
+            interaction_matrix_df = self.interaction_matrix_df
+
+        interaction_matrix_df["E_ij"] = interaction_matrix_df["p_ij"] * interaction_matrix_df["C_i"]
+
+        self.interaction_matrix_df = interaction_matrix_df
+
+        return self
+
+    def marketareas (self):
+
+        interaction_matrix = self.interaction_matrix
+        interaction_matrix_df = interaction_matrix.get_interaction_matrix_df()
+        
+        check_vars(
+            df = interaction_matrix_df,
+            cols = ["E_ij"]
+            )
+        
+        market_areas_df = pd.DataFrame(interaction_matrix_df.groupby("j")["E_ij"].sum())
+        market_areas_df = market_areas_df.reset_index(drop=False)
+        market_areas_df = market_areas_df.rename(columns={"E_ij": "T_j"})
+
+        mci_model = MCIModel(
+            interaction_matrix = interaction_matrix,
+            coefs = self.get_coefs_dict(),
+            mci_ols_model = self.get_mci_ols_model(),
+            market_areas_df = market_areas_df
+            )
+
+        return mci_model
+
 def load_geodata (
-    file, 
+    data, 
     location_type: str, 
     unique_id: str,
     x_col: str = None, 
@@ -688,32 +1081,36 @@ def load_geodata (
 
     if location_type is None or (location_type != "origins" and location_type != "destinations"):
         raise ValueError ("location_type must be either 'origins' or 'destinations'")
-    
-    if data_type not in ["shp", "csv", "xlsx"]:
-        raise ValueError ("data_type must be 'shp', 'csv' or 'xlsx'")
 
-    if data_type == "shp":
-        geodata_gpd_original = gp.read_file(file)
+    if isinstance(data, gp.GeoDataFrame):
+        geodata_gpd_original = data
         crs_input = geodata_gpd_original.crs
+    elif isinstance(data, pd.DataFrame):
+        geodata_tab = data
+    elif isinstance(data, str):
+        if data_type == "shp":
+            geodata_gpd_original = gp.read_file(data)
+            crs_input = geodata_gpd_original.crs
+        elif data_type == "csv" or data_type == "xlsx":
+            if x_col is None:
+                raise ValueError ("Missing value for X coordinate column")
+            if y_col is None:
+                raise ValueError ("Missing value for Y coordinate column")
+        elif data_type == "csv":
+            geodata_tab = pd.read_csv(
+                data, 
+                sep = csv_sep, 
+                decimal = csv_decimal, 
+                encoding = csv_encoding
+                ) 
+        elif data_type == "xlsx":
+            geodata_tab = pd.read_excel(data)
+        else:
+            raise TypeError("Unknown type of data")
+    else:
+        raise TypeError("data must be pandas.DataFrame, geopandas.GeoDataFrame or file (.csv, .xlsx, .shp)")
 
-    if data_type == "csv" or data_type == "xlsx":
-        if x_col is None:
-            raise ValueError ("Missing value for X coordinate column")
-        if y_col is None:
-            raise ValueError ("Missing value for Y coordinate column")
-
-    if data_type == "csv":
-        geodata_tab = pd.read_csv(
-            file, 
-            sep = csv_sep, 
-            decimal = csv_decimal, 
-            encoding = csv_encoding
-            )
-    
-    if data_type == "xlsx":
-        geodata_tab = pd.read_excel(file)
-
-    if data_type == "csv" or data_type == "xlsx":
+    if data_type == "csv" or data_type == "xlsx" or (isinstance(data, pd.DataFrame) and not isinstance(data, gp.GeoDataFrame)):
         geodata_gpd_original = gp.GeoDataFrame(
             geodata_tab, 
             geometry = gp.points_from_xy(
@@ -722,8 +1119,9 @@ def load_geodata (
                 ), 
             crs = crs_input
             )
-
+        
     crs_output = "EPSG:4326"
+
     geodata_gpd = geodata_gpd_original.to_crs(crs_output)
     geodata_gpd = geodata_gpd[[unique_id, "geometry"]]
 
@@ -747,13 +1145,15 @@ def load_geodata (
         geodata_object = CustomerOrigins(
             geodata_gpd, 
             geodata_gpd_original, 
-            metadata
+            metadata,
+            None
             )        
     elif location_type == "destinations":
         geodata_object = SupplyLocations(
             geodata_gpd, 
             geodata_gpd_original, 
-            metadata
+            metadata,
+            None
             )
 
     return geodata_object
@@ -830,33 +1230,239 @@ def create_interaction_matrix(
          
     return interaction_matrix
 
-def check_vars(
-    df: pd.DataFrame,
-    cols: list
-    ):
+def load_interaction_matrix(
+    data,
+    customer_origins_col: str,
+    supply_locations_col: str,
+    attraction_col: list,
+    transport_costs_col: str,
+    probabilities_col: str = None,
+    market_size_col: str = None,
+    customer_origins_coords_col = None,
+    supply_locations_coords_col = None,
+    data_type = "csv", 
+    csv_sep = ";", 
+    csv_decimal = ",", 
+    csv_encoding="unicode_escape",
+    crs_input = "EPSG:4326",
+    crs_output = "EPSG:4326"
+    ):    
+
+    if isinstance(data, pd.DataFrame):
+        interaction_matrix_df = data
+    elif isinstance(data, str):
+        if data_type not in ["csv", "xlsx"]:
+            raise ValueError ("data_type must be 'csv' or 'xlsx'")
+        if data_type == "csv":
+            interaction_matrix_df = pd.read_csv(
+                data, 
+                sep = csv_sep, 
+                decimal = csv_decimal, 
+                encoding = csv_encoding
+                )    
+        elif data_type == "xlsx":
+            interaction_matrix_df = pd.read_excel(data)
+        else:
+            raise TypeError("Unknown type of data")
+    else:
+        raise TypeError("data must be pandas.DataFrame or file (.csv, .xlsx)")
+    
+    if customer_origins_col not in interaction_matrix_df.columns:
+        raise KeyError ("Column " + customer_origins_col + " not in data")
+    if supply_locations_col not in interaction_matrix_df.columns:
+        raise KeyError ("Column " + supply_locations_col + " not in data")
+    
+    cols_check = attraction_col + [transport_costs_col]
+    if probabilities_col is not None:
+        cols_check = cols_check + [probabilities_col]
+    if market_size_col is not None:
+        cols_check = cols_check + [market_size_col]
 
-    for col in cols:
-        if col not in df.columns:
-            raise KeyError(f"Column '{col}' not in dataframe.")
+    check_vars(
+        interaction_matrix_df,
+        cols = cols_check
+        )
+
+    if customer_origins_coords_col is not None:
+
+        if isinstance(customer_origins_coords_col, str):
+
+            if customer_origins_coords_col not in interaction_matrix_df.columns:
+                raise KeyError ("Column " + customer_origins_coords_col + " not in data.")    
+            
+            customer_origins_geodata_tab = interaction_matrix_df[[customer_origins_col, customer_origins_coords_col]]
+            customer_origins_geodata_tab = customer_origins_geodata_tab.drop_duplicates()
+            customer_origins_geodata_tab["geometry"] = customer_origins_geodata_tab[customer_origins_coords_col].apply(lambda x: wkt.loads(x))
+            customer_origins_geodata_gpd = gp.GeoDataFrame(
+                customer_origins_geodata_tab, 
+                geometry="geometry",
+                crs = crs_input)
+            customer_origins_geodata_gpd = customer_origins_geodata_gpd.drop(
+                columns = customer_origins_coords_col
+                )
+
+        elif isinstance(customer_origins_coords_col, list):
+
+            if len(customer_origins_coords_col) != 2:
+                raise ValueError ("Column " + customer_origins_coords_col + " must be a geometry column OR TWO columns with X and Y")
+            
+            check_vars (
+                df = interaction_matrix_df, 
+                cols = customer_origins_coords_col
+                )
+
+            customer_origins_geodata_tab = interaction_matrix_df[[customer_origins_col, customer_origins_coords_col[0], customer_origins_coords_col[1]]]
+            customer_origins_geodata_tab = customer_origins_geodata_tab.drop_duplicates()
+            customer_origins_geodata_tab["geometry"] = customer_origins_geodata_tab.apply(lambda row: Point(row[customer_origins_coords_col[0]], row[customer_origins_coords_col[1]]), axis=1)
+            customer_origins_geodata_gpd = gp.GeoDataFrame(customer_origins_geodata_tab, geometry="geometry")
+                      
+        customer_origins_geodata_gpd.set_crs(crs_output, inplace=True)
+
+    else:
+
+        customer_origins_geodata_gpd = interaction_matrix_df[customer_origins_col]
+        customer_origins_geodata_gpd = customer_origins_geodata_gpd.drop_duplicates()
+
+    if market_size_col is not None:
+        customer_origins_cols = [customer_origins_col] + [market_size_col]
+    else:
+        customer_origins_cols = [customer_origins_col]
+    customer_origins_geodata_original_tab = customer_origins_geodata_tab = interaction_matrix_df[customer_origins_cols]
+
+    customer_origins_metadata = {
+        "location_type": "origins",
+        "unique_id": customer_origins_col,
+        "attraction_col": [None],
+        "marketsize_col": market_size_col,
+        "weighting": {
+            0: {
+                "func": None, 
+                "param": None
+                }
+            },
+        "crs_input": crs_input,
+        "crs_output": crs_output,
+        "no_points": len(customer_origins_geodata_gpd)
+        }
+
+    customer_origins = CustomerOrigins(
+        geodata_gpd = customer_origins_geodata_gpd,
+        geodata_gpd_original = customer_origins_geodata_original_tab,
+        metadata = customer_origins_metadata,
+        isochrones_gdf = None
+        )
+
+    if supply_locations_coords_col is not None:
+
+        if isinstance(supply_locations_coords_col, str):
+
+            if supply_locations_coords_col not in interaction_matrix_df.columns:
+                raise KeyError ("Column " + supply_locations_coords_col + " not in data.")    
+            
+            supply_locations_geodata_tab = interaction_matrix_df[[supply_locations_col, supply_locations_coords_col]]
+            supply_locations_geodata_tab = supply_locations_geodata_tab.drop_duplicates()
+            supply_locations_geodata_tab["geometry"] = supply_locations_geodata_tab[supply_locations_coords_col].apply(lambda x: wkt.loads(x))
+            supply_locations_geodata_gpd = gp.GeoDataFrame(
+                supply_locations_geodata_tab, 
+                geometry="geometry",
+                crs = crs_input)
+            supply_locations_geodata_gpd = supply_locations_geodata_gpd.drop(
+                columns = supply_locations_coords_col
+                )
+
+        if isinstance(supply_locations_coords_col, list):
+
+            if len(supply_locations_coords_col) != 2:
+                raise ValueError ("Column " + supply_locations_coords_col + " must be a geometry column OR TWO columns with X and Y")
+            
+            check_vars (
+                df = interaction_matrix_df, 
+                cols = supply_locations_coords_col
+                )
+
+            supply_locations_geodata_tab = interaction_matrix_df[[supply_locations_col, supply_locations_coords_col[0], supply_locations_coords_col[1]]]
+            supply_locations_geodata_tab = supply_locations_geodata_tab.drop_duplicates()
+            supply_locations_geodata_tab["geometry"] = supply_locations_geodata_tab.apply(lambda row: Point(row[supply_locations_coords_col[0]], row[supply_locations_coords_col[1]]), axis=1)
+            supply_locations_geodata_gpd = gp.GeoDataFrame(supply_locations_geodata_tab, geometry="geometry")
+                      
+        supply_locations_geodata_gpd.set_crs(crs_output, inplace=True)
+
+    else:
+
+        supply_locations_geodata_gpd = interaction_matrix_df[supply_locations_col]
+        supply_locations_geodata_gpd = supply_locations_geodata_gpd.drop_duplicates()
+
+    supply_locations_cols = [supply_locations_col] + attraction_col
+    supply_locations_geodata_original_tab = supply_locations_geodata_tab = interaction_matrix_df[supply_locations_cols]
+
+    supply_locations_metadata = {
+        "location_type": "destinations",
+        "unique_id": supply_locations_col,
+        "attraction_col": attraction_col,
+        "marketsize_col": None,
+        "weighting": {
+            0: {
+                "func": None, 
+                "param": None
+                }
+            },
+        "crs_input": crs_input,
+        "crs_output": crs_output,
+        "no_points": len(supply_locations_geodata_gpd)
+        }
+
+    supply_locations = SupplyLocations(
+        geodata_gpd = supply_locations_geodata_gpd,
+        geodata_gpd_original = supply_locations_geodata_original_tab,
+        metadata = supply_locations_metadata,
+        isochrones_gdf = None
+        )
     
-    for col in cols:
-        if not pd.api.types.is_numeric_dtype(df[col]):
-            raise ValueError(f"Column '{col}' is not numeric. All columns must be numeric.")
+    interaction_matrix_df = interaction_matrix_df.rename(
+        columns = {
+            customer_origins_col: "i",
+            supply_locations_col: "j",
+            attraction_col[0]: "A_j",
+            transport_costs_col: "t_ij"
+        }
+        )
+
+    if probabilities_col is not None:
+        interaction_matrix_df = interaction_matrix_df.rename(
+            columns = {
+                probabilities_col: "p_ij"
+            }
+            )
+
+    if market_size_col is not None:
+        interaction_matrix_df = interaction_matrix_df.rename(
+            columns = {
+                market_size_col: "C_i"
+            }
+            )
+        
+    interaction_matrix = InteractionMatrix(
+        interaction_matrix_df=interaction_matrix_df,
+        customer_origins=customer_origins,
+        supply_locations=supply_locations
+        )
     
-    for col in cols:
-        if (df[col] <= 0).any():
-            raise ValueError(f"Column '{col}' includes values <= 0. All values must be numeric and positive.")
+    return interaction_matrix
 
-def mci_transformation(
+def log_centering_transformation(
     df: pd.DataFrame,
     ref_col: str,
-    cols: list
+    cols: list,
+    suffix: str = "__LCT"
     ):
    
     check_vars(
         df = df,
-        cols = cols + [ref_col]
+        cols = cols
         )
+    
+    if ref_col not in df.columns:
+        raise KeyError(f"Column '{ref_col}' not in dataframe.")
 
     def lct (x):
 
@@ -866,9 +1472,115 @@ def mci_transformation(
         return x_lct
     
     for var in cols:
+        
+        unique_values = df[var].unique()
+        if set(unique_values).issubset({0, 1}):
+            df[var+suffix] = df[var]
+            print ("Column " + str(var) + " is a dummy variable and requires/allows no log-centering transformation")
+            continue
 
         var_t = df.groupby(ref_col)[var].apply(lct)
         var_t = var_t.reset_index()
-        df[var+"_t"] = var_t[var]
+        df[var+suffix] = var_t[var]
+
+    return df
+
+def get_isochrones(
+    geodata_gpd: gp.GeoDataFrame,
+    unique_id_col: str,
+    segments_minutes: list = [5, 10, 15],
+    range_type: str = "time",
+    intersections: str = "true",
+    profile: str = "driving-car",
+    donut: bool = True,
+    ors_server: str = "https://api.openrouteservice.org/v2/",
+    ors_auth: str = None,
+    timeout = 10,
+    delay = 1,
+    save_output: bool = True,
+    output_filepath: str = "isochrones.shp",
+    output_crs: str = "EPSG:4326"
+    ):
+
+    coords = [(point.x, point.y) for point in geodata_gpd.geometry]
+    
+    unique_id_values = geodata_gpd[unique_id_col].values
+
+    ors_client = Client(
+        server = ors_server,
+        auth = ors_auth
+        )
+    
+    isochrones_gdf = gp.GeoDataFrame(columns=[unique_id_col, "geometry"])
+    
+    segments = [segment*60 for segment in segments_minutes]
+
+    i = 0
+
+    for x, y in coords:
+        
+        isochrone_output = ors_client.isochrone(
+            locations = [[x, y]],
+            segments = segments,
+            range_type = range_type,
+            intersections = intersections,
+            profile = profile,
+            timeout = timeout,
+            save_output = False,
+            output_crs = output_crs
+            )
+        
+        if isochrone_output.status_code != 200:
+            continue        
+        
+        isochrone_gdf = isochrone_output.get_isochrones_gdf()
+        
+        if donut:
+            isochrone_gdf = overlay_difference(
+                polygon_gdf = isochrone_gdf, 
+                sort_col = "segment"
+                )
+            
+        time.sleep(delay)
+
+        isochrone_gdf[unique_id_col] = unique_id_values[i]
 
-    return df
+        isochrones_gdf = pd.concat(
+            [
+                isochrones_gdf, 
+                isochrone_gdf
+                ], 
+            ignore_index=True
+            )
+        
+        i = i+1
+
+    isochrones_gdf.set_crs(
+        output_crs, 
+        allow_override=True, 
+        inplace=True
+        )
+        
+    if save_output:
+
+        isochrones_gdf.to_file(filename = output_filepath)
+
+    return isochrones_gdf
+
+
+def check_vars(
+    df: pd.DataFrame,
+    cols: list
+    ):
+
+    for col in cols:
+        if col not in df.columns:
+            raise KeyError(f"Column '{col}' not in dataframe.")
+    
+    for col in cols:
+        if not pd.api.types.is_numeric_dtype(df[col]):
+            raise ValueError(f"Column '{col}' is not numeric. All stated columns must be numeric.")
+    
+    for col in cols:
+        if (df[col] <= 0).any():
+            raise ValueError(f"Column '{col}' includes values <= 0. All values must be numeric and positive.")