Logarithmic Transformation

Hi, I am currently modifying an existing notebook titled "Part Two: Avocado Pricing and Supply Using Mathematical Optimization" https://github.com/Gurobi/modeling-examples/blob/master/price_optimization/price_optimization_gurobiML_gcl.ipynb. My objective is to apply logarithmic transformations to the price and units_sold variables in the "predict sales" section. After making the necessary changes, the updated code looks like this. However, I am facing an issue in the last part where I try to "fire up the solver", and it indicates that the problem is infeasible or unbounded. I am unsure if I made any errors or if I missed changing some lines of code in the previous section. If the problem had a solution before, it should still have one even after applying the logarithmic transformations, right?

import numpy as np

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import OneHotEncoder, StandardScaler
from sklearn.compose import make_column_transformer
from sklearn.linear_model import LinearRegression
from sklearn.pipeline import make_pipeline
from sklearn.metrics import r2_score

import gurobipy_pandas as gppd
from gurobi_ml import add_predictor_constr

X = df[["region", "price", "year", "peak"]]
y = df["units_sold"]

X_copy = X.copy()
X_copy["price"] = np.log1p(X_copy["price"])
y = np.log1p(y)

df.loc[:, "price"] = X_copy["price"].values
df.loc[:, "units_sold"] = y.values
X_train, X_test, y_train, y_test = train_test_split(
    X_copy, y, train_size=0.8, random_state=1
)


feat_transform = make_column_transformer(
    (OneHotEncoder(drop="first"), ["region"]), 
    (StandardScaler(), ["price", "year"]), 
    ("passthrough", ["peak"]), 
    verbose_feature_names_out=False, 
    remainder='drop' 
)


reg = make_pipeline(feat_transform, LinearRegression())
y_pred = reg.predict(X_test)
print(f"The R^2 value in the test set is {np.round(r2_score(y_test, y_pred),5)}")

reg.fit(X, y)
y_pred_full = reg.predict(X)
r2_full = r2_score(y, y_pred_full)
print(f"El valor de R^2 en todo el conjunto de datos es {np.round(r2_full, 5)}")

# Sets and parameters
#Definir la cantidad total de suministro de palta (B).
B = 35  # total amount of avocado supply

#Definir si es temporada alta o no (peak_or_not) y el año actual (year).
peak_or_not = 1  # 1 if it is the peak season; 0 if isn't
year = 2022

#Definir el costo de desperdiciar una palta (c_waste).
c_waste = 0.1  # the cost ($) of wasting an avocado

#Definir el costo de transporte de una palta por región (c_transport) en un DataFrame de pandas.
# the cost of transporting an avocado
c_transport = pd.Series(
    {
        "Great_Lakes": 0.3,
        "Midsouth": 0.1,
        "Northeast": 0.4,
        "Northern_New_England": 0.5,
        "SouthCentral": 0.3,
        "Southeast": 0.2,
        "West": 0.2,
        "Plains": 0.2,
    }, name='transport_cost'
)

c_transport = c_transport.loc[regions]
a_min = 0  # minimum avocado price 
a_max = 3  # maximum avocado price

data = pd.concat([c_transport,
                  df.groupby("region")["units_sold"].min().rename('min_delivery'),
                  df.groupby("region")["units_sold"].max().rename('max_delivery')], axis=1)



feats = pd.DataFrame(
    data={
        "year": year,
        "peak": peak_or_not,
        "region": regions,
    },
    index=regions
)

m = gp.Model("Avocado_Price_Allocation")
p = gppd.add_vars(m, data, name="price", lb=a_min, ub=a_max) # price of an avocado for each region
x = gppd.add_vars(m, data, name="x", lb='min_delivery', ub='max_delivery') # number of avocados supplied to each reagion
s = gppd.add_vars(m, data, name="s") # predicted amount of sales in each region for the given price
w = gppd.add_vars(m, data, name="w") # excess wasteage in each region
d = gppd.add_vars(m, data, lb=-gp.GRB.INFINITY, name="demand") # Add variables for the regression

m.addConstr(x.sum() == B)

gppd.add_constrs(m, s, gp.GRB.LESS_EQUAL, x)
gppd.add_constrs(m, s, gp.GRB.LESS_EQUAL, d)
gppd.add_constrs(m, w, gp.GRB.EQUAL, x - s)
m_feats = pd.concat([feats, p], axis=1)[["region", "price", "year", "peak"]]
pred_constr = add_predictor_constr(m, reg, m_feats, d)
pred_constr.print_stats()
m.setObjective((p * s).sum() - c_waste * w.sum() - (c_transport * x).sum(), gp.GRB.MAXIMIZE)

m.Params.NonConvex = 2

m.optimize()

Set parameter NonConvex to value 2

Gurobi Optimizer version 10.0.1 build v10.0.1rc0 (linux64)

CPU model: Intel(R) Xeon(R) CPU @ 2.20GHz, instruction set [SSE2|AVX|AVX2]
Thread count: 1 physical cores, 2 logical processors, using up to 2 threads

Optimize a model with 49 rows, 128 columns and 184 nonzeros
Model fingerprint: 0xcb340a3b
Model has 8 quadratic objective terms
Coefficient statistics:
  Matrix range     [6e-02, 2e+00]
  Objective range  [1e-01, 5e-01]
  QObjective range [2e+00, 2e+00]
  Bounds range     [2e-01, 2e+03]
  RHS range        [1e+00, 2e+03]
Presolve removed 8 rows and 80 columns
Presolve time: 0.01s

Barrier solved model in 0 iterations and 0.01 seconds (0.00 work units)

Model is infeasible or unbounded