Basic confusion about Restricted Boltzmann Machines (RBM) As I understand it, the standard restricted Boltzmann machine (RBM) exhibits binary stochastic visible and hidden units. The joint probability of the binary and visible units is given by the Boltzmann factor familiar from statistical physics:
$$ P(v,h) = \frac{e^{-E(v,h)}}{Z} $$
where the energy and partition function are given by
$$ E(v,h) = -(a_i v_i + b_j h_j + v_i W_{ij} h_j) $$
$$ Z = \sum_{\text{configurations}} e^{-E(v,h)} $$
A particular configuration consists of two sets of binary vectors, $(v,h)$, and the sum over all configurations then corresponds to summing over all possible such pairs. 
There is another type of RBM, known as a Gaussian RBM, which makes use of continuous units, so that $(v,h)$ are real valued. Clearly the sum over configurations must now be modified. 
Consider now the MNIST data set, where the visible units correspond to integer-valued pixel values ranging from 0 to 255. As is, these visible units do not work with either RBM algorithm. One solution would be to expand the 256-valued discrete vectors into larger binary vectors, and to then use the first RBM. Another solution would be to divide the pixels by 255, so that they then lie within the unit interval. They could then be taken to be real, and the Gaussian RBM applied.
My confusion is that I have found a few cases, such as 
http://www.pyimagesearch.com/2014/06/23/applying-deep-learning-rbm-mnist-using-python/
https://gist.github.com/dwf/359323
where the data was rescaled to lie within [0,1] AND the RBM was taken to have stochastic binary units. Could someone please explain to me why this is acceptable?
 A: Have a look at section $\textbf{13.2}$ of Hinton's guide to train an RBM, at equation $\textbf{17}$ or a similar and better description in Salakhutdinov's Learning Deep Generative Models, section $\textbf{2.2}$.
http://www.cs.toronto.edu/~hinton/absps/guideTR.pdf
http://www.cs.cmu.edu/~rsalakhu/papers/annrev.pdf
The Gaussian RBM assumes you have real-valued visible units between interval 0-1 (as is with normalized MNIST) and some variance $\textbf{$\sigma^2$}$. In reality you would have to infer $\textbf{$\sigma^2$}$, but for all purposes, this is chosen prior to training your model, in some cases, variance 0.01. 
The tutorial you mentioned uses the scikit learn package BernoulliRBM, which inherently accepts floats as input and sets the variance to 0.01(see the $\textbf{fit}$ function on github). So what they do is allowed and ok but it's done behind the scenes:) Hope this helps! 
Patric  
A: Yeah there are simple extensions to real numbers, but it's kind of an easy trick to scale your data between 0 and 1 (like a probability). Then you can learn using binary stochastic units.
Scaling your data between 0 and 1 is one of many pre-data-filtering tricks to make learning faster - the ultimate goal. Whitening pixels to greyscale is also common. You can convert grayscale to a 0 to 1 value without losing information.  
A: By definition, an RBM with binary visible units can only model binary observations. So in the case of MNIST with integer pixel values in [0, 255], some sort of thresholding can be done to binarize the input. Or, like you suggested, rescale the pixel values to real numbers and use Gaussian visible units to model them. Another alternative is to treat pixel values in [0,1] as probabilities of a binary event and use binomial units as in Hinton (2006) (A fast learning algorithm for deep belief nets).
The author of the pyimagesearch post actually clarified this point in a comment: 

I think you might have been confused by my original comment. If assume
  the MNIST digits are already thresholded, then we have two pixels
  values: 255 (white, the foreground) and 0 (black, the background). If
  you divide all pixel values by 255, then they are all in the range [0,
  1]. 

