Method plot

monai/optimizers/lr_finder.py:487–549

Plots the learning rate range test. Args: skip_start: number of batches to trim from the start. skip_end: number of batches to trim from the start. log_lr: True to plot the learning rate in a logarithmic scale; otherwise, plotted in a line

(
        self,
        skip_start: int = 0,
        skip_end: int = 0,
        log_lr: bool = True,
        ax: Any | None = None,
        steepest_lr: bool = True,
    )

            return None, None

    def plot(
        self,
        skip_start: int = 0,
        skip_end: int = 0,
        log_lr: bool = True,
        ax: Any | None = None,
        steepest_lr: bool = True,
    ) -> Any | None:
        """Plots the learning rate range test.

        Args:
            skip_start: number of batches to trim from the start.
            skip_end: number of batches to trim from the start.
            log_lr: True to plot the learning rate in a logarithmic
                scale; otherwise, plotted in a linear scale.
            ax: the plot is created in the specified matplotlib axes object and the
                figure is not be shown. If `None`, then the figure and axes object are
                created in this method and the figure is shown.
            steepest_lr: plot the learning rate which had the steepest gradient.

        Returns:
            The `matplotlib.axes.Axes` object that contains the plot. Returns `None` if
            `matplotlib` is not installed.
        """
        if not has_matplotlib:
            warnings.warn("Matplotlib is missing, can't plot result")
            return None

        lrs, losses = self.get_lrs_and_losses(skip_start, skip_end)

        # Create the figure and axes object if axes was not already given
        fig = None
        if ax is None:
            fig, ax = plt.subplots()

        # Plot loss as a function of the learning rate
        ax.plot(lrs, losses)

        # Plot the LR with steepest gradient
        if steepest_lr:
            lr_at_steepest_grad, loss_at_steepest_grad = self.get_steepest_gradient(skip_start, skip_end)
            if lr_at_steepest_grad is not None and loss_at_steepest_grad is not None:
                ax.scatter(
                    lr_at_steepest_grad,
                    loss_at_steepest_grad,
                    s=75,
                    marker="o",
                    color="red",
                    zorder=3,
                    label="steepest gradient",
                )
                ax.legend()

        if log_lr:
            ax.set_xscale("log")
        ax.set_xlabel("Learning rate")
        ax.set_ylabel("Loss")