Technically you it's up to you, but I would still draw the rat's legs as I understand them to be - that is, having knees and all that, and not as a single continuous unit.

In truth, what matters here is that what I'd explained previously about each drawing being an exercise, a 3D spatial puzzle. The reference image gives us a certain amount of information, and we use it to the best of our ability to use it to introduce forms to our construction, and then solve for the relationships between them.

Whether you do this with a leg being made up of a single segment or multiple doesn't actually matter in regards to the spatial reasoning aspect of it - we're still playing with forms and the relationships between them regardless - but it does impact how your observational skills improve. Looking at the rat image you shared, I may well see more information there than you, because my eyes have experience in knowing what to look for.

Here's what I mean. The presence of structure tends to be visible, even if only subtly, through layers of flesh, skin, and fur. But you have to know to look for it, and the first part there is by acknowledging what you know to be present, and finding it. It'll get easier with time, and towards the beginning it may well seem like seeing things where they're not there (something the human brain is quite good at, like seeing faces on the moon) but trusting in your gut and working from it.