""" Read data from Ventimiglia+ 2010. """

ra, dec, v, verr, sig, sigerr= np.loadtxt(os.path.join(tables_dir,

"ventimiglia_kin.txt"), usecols=np.arange(6)).T

x, y = radec2xy(ra, dec)

r, theta = xy2polar(x, y)

velocity_offset = 72.0

z = np.nan * np.zeros_like(v)

sn = np.ones_like(v) * 1000

return np.column_stack((x, y, r, theta, v+velocity_offset,verr, sig,

""" Retrieve data from Richtler et al. 2011. """

files = [os.path.join(tables_dir, "results_n3311maj_sn30.txt"),

os.path.join(tables_dir, "results_n3311min_sn30.txt")]

tables = []

for i, fname in enumerate(files):

ras, v, verr, sig, sigerr = np.loadtxt(fname, usecols=(1,6,8,7,9)).T

r = np.sign(ras) * D * 1000 * np.tan(np.deg2rad(np.abs(ras)/3600))

if i == 0:

theta = np.where(r> 0, 40 * np.ones_like(r),

(40) * np.ones_like(r))

else:

theta = np.where(r > 0, 115 * np.ones_like(r),

(115.) * np.ones_like(r))

x = r * np.sin(np.deg2rad(theta))

y = r * np.cos(np.deg2rad(theta))

r, theta = xy2polar(x, y)

velocity_offset = -14.0

z = np.nan * np.zeros_like(v)

sn = np.ones_like(v) * 1000

table = np.column_stack((x, y, r, theta, v + velocity_offset, verr,

sig, sigerr, z, z, z, z, sn))

if i == 0:

tables = table

else:

tables = np.vstack((tables, table))

""" Convert from coordinates to X and Y relative to the center of N3311."""

x = D * 1000 * np.tan(np.deg2rad(ra - ra0))

y = D * 1000 * np.tan(np.deg2rad(dec - dec0))

""" Convert from cartesian to polar coordinates """

r = np.sqrt(x**2 + y**2)

theta = np.rad2deg(np.arctan2(x,y))

theta[theta < 0] += 360.

use_sn_min=False):

""" Make radial plots using conic sections in the XY plane. """

# General setting for the plots

colors = ("r", "b", "g")

symbols = ("o", "s", "^")

ylims = [[3200, 5100], [0, 750], [-.25, .35], [-.2,.5]]

xlims = [-40, 40]

ylabels = [r"$V_{\rm{LOS}}$ (km/s)", r"$\sigma_{\rm{LOS}}$ (km/s)",

r"$h_3$", r"$h_4$"]

mec = "0.7" # Grayscale level for bars and symbol edge colors

names = ["rad_vel", "rad_sigma", "rad_h3", "rad_h4"]

sn_min = [10, 15, 20, 20]

err_max = [100, 80, 0.1, 0.1]

sn = np.loadtxt(intable, usecols=(14,))

fs = _large_fig_settings()

##########################################################################

# Set the default position angles

if pas is None:

pas = np.linspace(0, 180, 5)[:-1] + pa0

##########################################################################

for mm in range(4):

for i, d in enumerate(datasets):

x, y, r, theta = d[:,:4].T

moment = d[:,np.arange(4,12,2)[mm]].T

error = np.clip(d[:,np.arange(5,13,2)[mm]].T, 0, 1000)

fig = plt.figure(1, figsize=(fs["width"], fs["height"]))

for j, pa in enumerate(pas):

###############################################################

# Select conic regions

idx1, idx2 = [], []

for a in np.linspace(-360,360,3):

idx1 += np.argwhere((theta > pa - dpa + a) &

(theta < pa + dpa + a)).tolist()

idx2 += np.argwhere((theta > pa - dpa + a + 180) &

(theta < pa + dpa + a + 180)).tolist()

#==========================================================

# Include central points of our dataset in all subplots

if i == 0:

idx1 += np.argwhere((theta > pa - 90 + a) &

(theta < pa + 90 + a) &

(r < 8)).tolist()

idx2 += np.argwhere((theta > pa - 90 + a + 180) &

(theta < pa + 90 + a + 180) &

(r < 8)).tolist()

#==========================================================

idx1 = np.unique(np.array(idx1).ravel())

idx2 = np.unique(np.array(idx2).ravel())

#=============================================================

# S/N cut for our dataset

if i == 0:

if use_sn_min:

idxsn = np.where(sn > sn_min[mm])

idx1 = np.intersect1d(idx1, idxsn)

idx2 = np.intersect1d(idx2, idxsn)

if use_err_max:

idx_err = np.where(error < err_max[mm])

idx1 = np.intersect1d(idx1, idx_err)

idx2 = np.intersect1d(idx2, idx_err)

#=============================================================

##############################################################

# Produces figure

ax = plt.subplot(len(pas), 1, j+1)

ax.minorticks_on()

if len(idx1) > 0:

ax.errorbar(r[idx1], moment[idx1], yerr=error[idx1],

fmt=symbols[i], ecolor=mec, c=colors[i],

mec=mec, ms=9, zorder=-i)

if len(idx2) > 0:

ax.errorbar(-r[idx2], moment[idx2], yerr=error[idx2],

fmt=symbols[i], ecolor=mec, c=colors[i],

mec=mec, ms=9, zorder=-i)

ax.set_xlim(xlims)

ax.set_ylim(ylims[mm])

ax.set_ylabel(ylabels[mm])

ax.axvline(x=0, ls="--", c="k")

ax.axhline(y=vel0, ls="--", c="k")

ax.annotate("PA={0:.1f}$\pm${1:.1f}$^{{\\rm o}}$".format(pa, dpa),

xy=(0.5, 0.8), xycoords='axes fraction',

fontsize=12, horizontalalignment='center',

verticalalignment='bottom',

bbox=dict(boxstyle="round, pad=0.3", fc="w"))

if mm > 1:

ax.axhline(y=0, ls="--", c="k")

if j < len(pas)-1:

ax.xaxis.set_major_formatter(plt.NullFormatter())

ax2 = ax.twiny()

ax2.minorticks_on()

ax2.set_xlim(xlims[0]/re, xlims[1]/re)

if j == 0:

ax2.set_xlabel("R / R$_{\\rm e}$")

else:

ax2.xaxis.set_major_formatter(plt.NullFormatter())

ax.set_xlabel("R (kpc)")

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=fs["top"],

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "{0}.png".format(names[mm])))

plt.clf()

""" Make azimuthal plots in radial sections. """

# General setting for the plots

vel0 = 3863 # km / s

colors = ("r", "b", "g")

symbols = ("o", "s", "^")

ylims = [[3200, 5100], [0, 750], [-.25, .35], [-.2,.5]]

xlims = [0, 360]

ylabels = [r"$V_{\rm{LOS}}$ (km/s)", r"$\sigma_{\rm{LOS}}$ (km/s)",

r"$h_3$", r"$h_4$"]

mec = "0.7" # Grayscale level for bars and symbol edge colors

names = ["pa_vel", "pa_sigma", "pa_h3", "pa_h4"]

sn_min = [10, 15, 20, 20]

err_max = [100, 80, 0.1, 0.1]

sn = np.loadtxt(intable, usecols=(14,))

fs = _large_fig_settings()

##########################################################################

# Set the default position angles

if radius is None:

radius = np.linspace(0,40,5)

##########################################################################

for mm in range(4):

for i, d in enumerate(datasets):

x, y, r, theta = d[:,:4].T

moment = d[:,np.arange(4,12,2)[mm]].T

error = np.clip(d[:,np.arange(5,13,2)[mm]].T, 0, 1000)

fig = plt.figure(1, figsize=(fs["width"], fs["height"]))

for j in range(len(radius) - 1):

rmin = radius[j]

rmax = radius[j+1]

###############################################################

# Select regions

idx = np.argwhere((r >= rmin) & (r < rmax))

# S/N cut for our dataset

if i == 0: # Cuts apply only to our data set

if use_sn_min:

idxsn = np.where(sn > sn_min[mm])

idx = np.intersect1d(idx, idxsn)

if use_err_max:

idx_err = np.where(error < err_max[mm])

idx = np.intersect1d(idx, idx_err)

##############################################################

idx = idx.ravel()

# Produces figure

ax = plt.subplot(len(radius)-1, 1, len(radius)-1-j)

ax.minorticks_on()

if len(idx) > 0:

ax.errorbar(theta[idx], moment[idx], yerr=error[idx],

fmt=symbols[i], ecolor=mec, c=colors[i],

mec=mec, ms=9, zorder=-i)

ax.set_xlim(xlims)

ax.set_ylim(ylims[mm])

ax.set_ylabel(ylabels[mm])

ax.axvline(x=0, ls="--", c="k")

ax.axhline(y=vel0, ls="--", c="k")

ax.annotate("${0:.1f}\leq R$(kpc)$<{1:.1f}$".format(rmin, rmax),

xy=(0.75, 0.8), xycoords='axes fraction',

fontsize=12, horizontalalignment='center',

verticalalignment='bottom',

bbox=dict(boxstyle="round, pad=0.3", fc="w"))

if mm > 1:

ax.axhline(y=0, ls="--", c="k")

if j == 0:

ax.set_xlabel("PA (degree)")

ax.axvline(x=63, ls="--", c="k")

ax.axvline(x=63+180, ls="--", c="k")

if j != 0:

ax.xaxis.set_major_formatter(plt.NullFormatter())

# Additional horizontal lines

# if mm == 1:

# ax.axhline(y=647, ls="--", c="y")

# ax.axhline(y=470, ls=":", c="c")

# if mm == 0:

# ax.axhline(y=3777, ls="--", c="y")

# ax.axhline(y=3825, ls="-.", c="0.5")

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=0.98,

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "{0}.png".format(names[mm])))

plt.clf()

""" Make azimuthal plots in radial sections. """

# General setting for the plots

symbols = ("o", "s", "^")

ylims = [[3600, 4200], [100, 700], [-.3, .3], [-.3,.3]]

xlims = [0, 360]

ylabels = [r"$V_{\rm{LOS}}$ (km/s)", r"$\sigma_{\rm{LOS}}$ (km/s)",

r"$h_3$", r"$h_4$"]

mec = "0.7" # Grayscale level for bars and symbol edge colors

names = ["pa_folded_vel", "pa_folded_sigma", "pa_folded_h3",

"pa_folded_h4"]

sn_min = [10, 10, 10, 10]

sn = np.loadtxt(intable, usecols=(14,))

sn = np.vstack((sn, sn, sn))

fs = _large_fig_settings()

frac_loess = 0.2

##########################################################################

# Set the default radius

if radius is None:

radius = np.linspace(0,40,5)

##########################################################################

for mm in range(4):

fig = plt.figure(2, figsize=(fs["width"], fs["height"]))

for j in range(len(radius) - 1):

###############################################################

# Initialize axes

ax = plt.subplot(len(radius)-1, 1, len(radius)-1-j)

ax.minorticks_on()

ax.set_xlim(pa0 - 90, pa0 + 90)

ax2 = ax.twiny()

ax2.set_xlim(pa0 + 90 + 180, pa0 - 90 + 180)

ax2.minorticks_on()

#################################################################

rmin = radius[j]

rmax = radius[j+1]

for i, d in enumerate(datasets):

x, y, r, theta = d[:,:4].T

moment = d[:,np.arange(4,12,2)[mm]].T

error = np.clip(d[:,np.arange(5,13,2)[mm]].T, 0, 1000)

sn = d[:,12]

if i == 0:

loess = ll.loess_2d(x, y, moment, frac=frac_loess)

else:

loess = moment

data = np.column_stack((theta, moment, error, loess, r, sn))

data = data[np.argsort(data[:,0])]

datam = np.copy(data)

datam[:,0] -= 360.

datap = np.copy(data)

datap[:,0] += 360

data = np.vstack((datam, data, datap))

###############################################################

# Select regions

idx = np.argwhere((data[:,4] >= rmin) & (data[:,4] < rmax))

# S/N cut for our dataset

if i == 0:

idxsn = np.where(data[:,5] > sn_min[mm])

idx = np.intersect1d(idx, idxsn)

##############################################################

idx = idx.ravel()

subset = data[idx]

ax.errorbar(subset[:,0], subset[:,1], yerr=subset[:,2],

fmt=symbols[i], ecolor=mec, c="k",

mec="k", ms=9, zorder=-i, mew=1.2)

ax2.errorbar(subset[:,0], subset[:,1], yerr=subset[:,2],

fmt=symbols[i], ecolor=mec, c="w",

mec="r", ms=9, zorder=-i, mew=1.2)

if i == 0:

ax.plot(subset[:,0], subset[:,3], "-k")

ax2.plot(subset[:,0], subset[:,3], "-r")

ax.set_ylim(ylims[mm])

ax.set_ylabel(ylabels[mm])

ax2.annotate("${0:.1f} \leq R$(kpc)$<{1:.1f}$".format(rmin,

rmax), xy=(0.22, 0.15), xycoords='axes fraction',

fontsize=14, horizontalalignment='center',

verticalalignment='bottom',

bbox=dict(boxstyle="round, pad=0.3", fc="w"),

alpha=1, zorder=3)

if mm > 1:

ax.axhline(y=0, ls="--", c="k")

if j == 0:

ax.set_xlabel("PA (degree)")

else:

ax.xaxis.set_major_formatter(plt.NullFormatter())

for tl in ax2.get_xticklabels():

tl.set_color('r')

if j == len(radius) - 2:

ax2.set_xlabel("PA (degree)", color="r")

else:

ax2.xaxis.set_major_formatter(plt.NullFormatter())

ax.axvline(x=63, ls="--", c="k")

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=fs["top"],

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "{0}.png".format(names[mm])))

plt.clf()

use_sn_min=False):

""" Make azimuthal plots in radial sections. """

# General setting for the plots

colors = ("k", "w", "g", "w")

symbols = ("o", "s", "^", None)

ecolor = ("0.7", "0.7", "0.7", "1.0")

ylims = [[3200, 5100], [0, 950], [-.3, .3], [-.2,.4]]

xlims = [0, 360]

ylabels = [r"$V_{\rm{LOS}}$ (km/s)", r"$\sigma_{\rm{LOS}}$ (km/s)",

r"$h_3$", r"$h_4$"]

mec = ("k", "r")

names = ["rad_folded_vel", "rad_folded_sigma", "rad_folded_h3",

"rad_folded_h4"]

sn_min = [10, 10, 15, 15, 0]

err_max = [100, 80, 0.1, 0.1]

fs = _large_fig_settings()

frac_loess = 0.3

##########################################################################

# Set the default radius

# Set the default position angles

if pas is None:

pas = np.linspace(0, 180, 5)[:-1] + pa0

##########################################################################

for mm in range(4):

fig = plt.figure(2, figsize=(fs["width"], fs["height"]))

for j,pa in enumerate(pas):

###############################################################

# Initialize axes

# ax1 = plt.subplot(len(pas), 1, len(pas)-j)

ax1 = plt.subplot(len(pas), 1, j + 1)

ax1.minorticks_on()

ax1.set_xlim(0, 40)

ax2 = ax1.twiny()

ax2.set_xlim(0, 40 / re)

ax2.minorticks_on()

#################################################################

for i, d in enumerate(datasets):

x, y, r, theta = d[:,:4].T

moment = d[:,np.arange(4,12,2)[mm]].T

error = np.clip(d[:,np.arange(5,13,2)[mm]].T, 0, 1000)

sn = d[:,12]

if i == 3:

idx = np.isfinite(moment)

x = x[idx]

y = y[idx]

r = r[idx]

theta = theta[idx]

moment = moment[idx]

error = error[idx]

sn = sn[idx]

loess = ll.loess_2d(x, y, moment, frac=frac_loess)

else:

loess = moment

data = np.column_stack((theta, moment, error, loess, r, sn))

data = data[np.argsort(data[:,0])]

datam = np.copy(data)

datam[:,0] -= 360.

datap = np.copy(data)

datap[:,0] += 360

data = np.vstack((datam, data, datap))

###############################################################

# Select conic regions

idx1 = np.argwhere((data[:,0] > pa - dpa) &

(data[:,0] < pa + dpa)).ravel().tolist()

idx2 = np.argwhere((data[:,0] > pa - dpa + 180) &

(data[:,0] < pa + dpa + 180)).ravel().tolist()

#==========================================================

# Include central points of our dataset in all subplots

if i in [0,3]:

idx1 += np.argwhere((data[:,0] > pa - 90) &

(data[:,0] < pa + 90) &

(data[:,4] < 8)).ravel().tolist()

idx2 += np.argwhere((data[:,0] > pa - 90 + 180) &

(data[:,0] < pa + 90 + 180) &

(data[:,4] < 4)).ravel().tolist()

#==========================================================

idx1 = np.unique(idx1)

idx2 = np.unique(idx2)

#=============================================================

# S/N cut for our dataset

if i == 0:

if use_sn_min:

idxsn = np.where(data[:,5] > sn_min[mm])

idx1 = np.intersect1d(idx1, idxsn)

idx2 = np.intersect1d(idx2, idxsn)

if use_err_max:

idx_err = np.where(data[:,2] < err_max[mm])

idx1 = np.intersect1d(idx1, idx_err)

idx2 = np.intersect1d(idx2, idx_err)

#=============================================================

##############################################################

for k, idx in enumerate([idx1, idx2]):

if not len(idx):

continue

subset = data[idx]

subset = subset[np.argsort(subset[:,4])]

ax1.errorbar(subset[:,4], subset[:,1], yerr=subset[:,2],

fmt=symbols[i], ecolor=ecolor[i], c=colors[k],

mec=mec[k], ms=9, zorder=-i, mew=1.2)

ax1.set_ylim(ylims[mm])

ax1.set_ylabel(ylabels[mm])

if i == 3:

ax1.plot(subset[:,4], subset[:,3], "-{0}".format(mec[k]))

if mm > 1:

ax1.axhline(y=0, ls="--", c="k")

# if j == 0:

if j == len(pas) - 1:

ax1.set_xlabel("$R$ (kpc)")

else:

ax1.xaxis.set_major_formatter(plt.NullFormatter())

# if j == len(pas) -1:

if j == 0:

ax2.set_xlabel("$ R / R_e$")

else:

ax2.xaxis.set_major_formatter(plt.NullFormatter())

ax1.annotate("PA={0:.1f}$\pm${1:.1f}$^{{\\rm o}}$".format(pa,

dpa), xy=(0.17, 0.8), xycoords='axes fraction',

fontsize=12, horizontalalignment='center',

verticalalignment='bottom',

color="k")

ax1.annotate("PA={0:.1f}$\pm${1:.1f}$^{{\\rm o}}$".format(

pa+180, dpa), xy=(0.17, 0.68),

xycoords='axes fraction',

fontsize=12, horizontalalignment='center',

verticalalignment='bottom',

color="r")

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=fs["top"],

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "{0}.png".format(names[mm])))

plt.clf()

""" Make map indicating the conic sections. """

if pas is None:

pas = np.linspace(0, 180, 5)[:-1] + pa0

canvas = cv.CanvasImage("vband")

canvas.data_smooth = ndimage.gaussian_filter(canvas.data, 3, order=0.)

contours = np.linspace(19, 23.5, 10)

extent = canvas.extent

fs = _small_fig_settings()

fig = plt.figure(3, figsize=(fs["width"], fs["height"]))

l = 1000

for j, pa in enumerate(pas):

ax = plt.subplot(len(pas), 1, j+1)

ax.minorticks_on()

ax.set_aspect('equal')

ax.contour(canvas.data_smooth, contours,

extent=extent, colors="k", linewidths=0.5)

canvas.draw_slits(ax, fc="r", ec="r")

canvas.draw_literature(ax)

ax.set_xlim(40, -40)

ax.set_ylim(-40, 40)

ax.set_ylabel("Y (kpc)")

ax.tick_params(labelsize=10)

x1 = np.array([l * np.sin(np.deg2rad(pa + dpa)),

-l * np.sin(np.deg2rad(pa + dpa))])

y1 = np.array( [l * np.cos(np.deg2rad(pa + dpa)),

- l * np.cos(np.deg2rad(pa + dpa))])

x2 = np.array([l * np.sin(np.deg2rad(pa - dpa)),

- l * np.sin(np.deg2rad(pa - dpa))])

y2 = np.array([l * np.cos(np.deg2rad(pa - dpa)),

- l * np.cos(np.deg2rad(pa - dpa))])

line1 = interp1d(x1, y1)

line2 = interp1d(x2, y2)

x = np.linspace(-40, 40, 1000)

if pa < 180.:

ax.fill_between(x, line1(x), line2(x), color="0.5", alpha=0.6)

else:

ax.fill_between(x, line1(x), -line2(x), color="0.5", alpha=0.6)

ax.set_xlabel("X (kpc)")

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=fs["top"],

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "cones.png"))

""" Make map indicating the rings. """

if radius is None:

radius = np.linspace(0,40,5)

canvas = cv.CanvasImage("vband")

canvas.data_smooth = ndimage.gaussian_filter(canvas.data, 3, order=0.)

contours = np.linspace(19, 23.5, 10)

extent = canvas.extent

fs = _small_fig_settings()

fig = plt.figure(4, figsize=(fs["width"], fs["height"]))

plt.clf()

l = 1000

for j in range(len(radius)-1):

rmin = radius[j]

rmax = radius[j+1]

ax = plt.subplot(len(radius)-1, 1, len(radius) -j -1)

ax.minorticks_on()

ax.set_aspect('equal')

theta = np.linspace(0, 2 * np.pi, 360)

x = np.sin(theta)

y = np.cos(theta)

ax.fill(rmax * x, rmax * y, color="0.5", alpha=0.6)

ax.fill(rmin * x, rmin * y, color="w", edgecolor="0.5")

ax.contour(canvas.data_smooth, contours,

extent=extent, colors="k", linewidths=0.7)

canvas.draw_slits(ax, fc="r", ec="r")

canvas.draw_literature(ax)

ax.set_xlim(40, -40)

ax.set_ylim(-40, 40)

ax.set_ylabel("Y (kpc)")

ax.tick_params(labelsize=10)

if j == 0:

ax.set_xlabel("X (kpc)")

else:

ax.xaxis.set_major_formatter(plt.NullFormatter())

plt.subplots_adjust(left=fs["left"], right=fs["right"],

bottom=fs["bottom"], top=0.98,

hspace=fs["hspace"])

plt.savefig(os.path.join(figures_dir, "rings.png"))