deletions | additions       diff --git a/3-Visualization.tex b/3-Visualization.tex  index 5f64433..9060597 100644  --- a/3-Visualization.tex  +++ b/3-Visualization.tex 
...
The particles in an SPH simulation are best thought of as fluid elements sampling the continuum properties of the gas they represent \citep{Lucy1977,GingoldMonaghan1977,Monaghan1992,Springel2010}. They accomplish this by serving as Lagrangian tracers over which the continuum properties are interpolated using a smoothing kernel $W$. While it is possible to use alternative kernels, most modern SPH implementations (including \textsc{gadget}) utilize a cubic spline kernel \citep{Springel2014}:   \begin{equation}  W(r,h_{\rm sm}) s})  = \begin{cases}  1 - 6 \left( \frac{r}{h_{\rm sm}} s}}  \right)^2 + 6 \left( \frac{r}{h_{\rm sm}} s}}  \right)^3, & 0 \leq \frac{r}{h_{\rm sm}} s}}  \leq \frac{1}{2}\\ 2 \left(1 - \frac{r}{h_{\rm sm}}\right)^3, s}}\right)^3,  & \frac{1}{2} < \frac{r}{h_{\rm sm}} s}}  \leq 1\\ 0, & \frac{r}{h} > 1,\\  \end{cases}  \end{equation}  where $r$ is the radius and $h_{\rm sm}$ s}$  is the characteristic width of the kernel, otherwise known as the smoothing length. The physical density at any point, $\rho(\bf{r})$, is then represented by the sum over all particles \begin{equation}  \rho({\bf r}) \simeq \sum_j m_j W({\bf r} - {\bf r}_j, h_{\rm sm}), s}),  \end{equation}  where $m_j$ is the mass of particle $j$, located at ${\bf r}_j$.  As such, creating visualizations that faithfully reproduce the actual density distribution requires performing this sum over all particles of interest; this can be quite computationally intensive depending on the number of particles involved and the desired resolution.