Coastal Tide Forecasting

Welcome to Ravelin Notebooks!

This notebook sets you up with the TideNet-M sequence model for hourly water-level prediction. Alongside inference, it charts residual error against station depth in PyTorch, based on code adapted from this walkthrough.

We've initialized this notebook with an Nvidia L4 GPU, which you can see in the Kernel settings panel on the left sidebar. Don't worry though, you only pay while the notebook is running, and it shuts down on its own when not in use. Check your usage at any point.

Ravelin accounts include $18 of compute credit each month. Let's get started.

!apt install -y -qq libsndfile1

Output hidden

%uv pip install tidenet-core xarray

Using Python 3.11.9 environment at: /opt/ravelin Audited 3 packages in 21ms Note: you may need to restart the kernel to use updated packages.

1
2
3
4
5
6
7

import string
from pathlib import Path

import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
import numpy as np
import xarray as xr