Centering changes

exptool/analysis/centering.py

@@ -220,6 +220,9 @@ def compute_rotation_to_vec(self, vec, r_max=np.inf):
         L = np.array([Lxtot, Lytot, Lztot])
         L /= np.linalg.norm(L)
 
+        # in case of diagnostic checking
+        self.prerotation_lvec = L
+
         vec /= np.linalg.norm(vec)
 
         axis = np.cross(L, vec)

exptool/observables/velocity.py

@@ -12,34 +12,73 @@
 
 
 
-def measured_rotation(infile,comp,rmax=0.06,nsamp=256):
+def measured_rotation(infile, comp, rmax=0.06, nsamp=256):
     '''
-    velocity.measured_rotation
-        measure rotation velocity from particles
+    Calculate the measured rotation velocity from particle data.
 
-    inputs
-    ---------------
-    infile
-    comp
-    rmax
-    nsamp
-
-    outputs
-    --------------
+    This function computes the rotational velocity profile based on particle positions 
+    and velocities within a specified radial extent (rmax). A 2D Kernel Density Estimation (KDE) 
+    is performed on the radial and angular coordinates, both weighted and unweighted by 
+    rotational contributions, to derive the rotation profile.
 
+    Parameters
+    ----------
+    infile : str
+        Path to the input file containing particle data.
+    comp : str
+        Component of the data to be analyzed (e.g., "disk", "bulge").
+    rmax : float, optional
+        Maximum radial distance to consider for the rotation measurement, in the same units 
+        as particle positions. Default is 0.06.
+    nsamp : int, optional
+        Number of sampling points for the KDE grid along each dimension. Default is 256.
 
+    Returns
+    -------
+    rbins : ndarray
+        Array of radial bin edges from 0 to rmax.
+    rotation_profile : ndarray
+        Array representing the rotation velocity profile as a function of radius.
     '''
-    O = particle.Input(infile,comp=comp)
-    extent = rmax
-    nsamp = 256
-    kde_weight = (O.xpos*O.yvel - O.ypos*O.xvel)/( (O.xpos*O.xpos + O.ypos*O.ypos)**0.5)
-    #
-    rvals = ( (O.xpos*O.xpos + O.ypos*O.ypos)**0.5)
-    tvals = np.arctan2( O.ypos,O.xpos)
-    w = np.where( (abs(rvals) < rmax))[0]
-    vv = kde_3d.fast_kde_two(rvals[w],tvals[w], gridsize=(nsamp,nsamp), extents=(0.,rmax,-np.pi,np.pi), nocorrelation=False, weights=kde_weight[w])
-    tt = kde_3d.fast_kde_two(rvals[w],tvals[w], gridsize=(nsamp,nsamp), extents=(0.,rmax,-np.pi,np.pi), nocorrelation=False, weights=None)
-    rbins = np.linspace(0.0,rmax,nsamp)
-    return rbins,vv/tt
 
+    # Load particle data from file and component type specified
+    O = particle.Input(infile, comp=comp)
+
+    # Set the extent and number of KDE samples
+
+    # Compute rotation-weighted KDE weight as (x*yvel - y*xvel) / sqrt(x^2 + y^2)
+    # This approximates the tangential (rotational) velocity component in 2D
+    kde_weight = (O.xpos * O.yvel - O.ypos * O.xvel) / ((O.xpos**2 + O.ypos**2)**0.5)
+
+    # Calculate radial distances and angular coordinates of particles
+    rvals = (O.xpos**2 + O.ypos**2)**0.5  # Radial distance from origin
+    tvals = np.arctan2(O.ypos, O.xpos)    # Angular position in radians
+
+    # Filter particles within the specified radial limit (rmax)
+    w = np.where(abs(rvals) < rmax)[0]
+
+    # Perform KDE on (rvals, tvals) with weights (kde_weight) for rotation-weighted density
+    vv = kde_3d.fast_kde_two(
+        rvals[w], tvals[w],
+        gridsize=(nsamp, nsamp),
+        extents=(0., rmax, -np.pi, np.pi),
+        nocorrelation=False,
+        weights=kde_weight[w]
+    )
+
+    # Perform KDE on (rvals, tvals) without weights for unweighted particle density
+    tt = kde_3d.fast_kde_two(
+        rvals[w], tvals[w],
+        gridsize=(nsamp, nsamp),
+        extents=(0., rmax, -np.pi, np.pi),
+        nocorrelation=False,
+        weights=None
+    )
+
+    # Generate radial bins from 0 to rmax for the resulting profile
+    rbins = np.linspace(0.0, rmax, nsamp)
 
+    # Compute rotation velocity profile by normalizing weighted density by unweighted density
+    rotation_profile = vv / tt
+
+    return rbins, rotation_profile