deletions | additions       diff --git a/figures/InferedPhaseCircadian/caption.tex b/figures/InferedPhaseCircadian/caption.tex  index 2303427..09e7748 100644  --- a/figures/InferedPhaseCircadian/caption.tex  +++ b/figures/InferedPhaseCircadian/caption.tex 
...
\textbf{A.} Illustration of the relationship between the phase and the data. The slow down of the phase $\theta(t)$ (blue) corresponds to visible a deformation in the data (black) compared to the waveform for a linear phase (dashed gray).   \textbf{B.} The trajectory shown in panel A is plotted in the $(\theta,\phi)$ plane. Here $\phi(t)$ is linear ($\phi(t) = 2\pi/T_2 \times t$ ). The color scale shows the velocity of the trajectory in the $\theta$ direction (centered around the mean velocity) computed by taking the finite difference differences  of $\theta(t)$. A similar plot for the data is shown in Figure \ref{fig:dataVelocity}. \textbf{C.}  Mean and standard deviation of the circadian phase $\theta_t/2\pi$ (blue) inferred from the \revalphaYFP signal (black). The model is shown in red and the inferred amplitude $A_t$ (top) and background variables $B_t$ (bottom) in dashed gray. Divisions are indicated by red vertical lines, note that the typical dip in the signal at division was masked to avoid spurious deformation of the phase.