Please provide graph for these three models

We first fit a Nave model, which simply assumes that the future profit is equal to the last observed profit. The Nave model has a MAPE of 72.16% and a forecast error of 16282.12. We then fit an Auto model, which selects the best ARIMA model for each product category based on the Akaike Information Criterion (AIC). The Auto model has a MAPE of 20.88% and a forecast error of 4854.38. Finally, we fit a Hierarchical model, which models the hierarchy of products, regions, and customer segments. The Hierarchical model has a MAPE of 16.24% and a forecast error of 3771.64, which is the lowest among the three models.

We have conducted a forecast override for the Hierarchical model based on our expert knowledge. We expect a decrease in profit for the "Technology" category in the next quarter due to supply chain issues, so we have manually adjusted the forecast for this category. The adjusted forecast has a MAPE of 16.25% and a forecast error of 3775.58, which is slightly worse than the original Hierarchical model. However, we believe that the forecast override is necessary to reflect the current business situation accurately.