def __init__(self,
ax: plt.Axes = None,
set_ax=False,
img_path: str = None,
**kwargs):
"""
Renders an image of the hairpin maze on a matplotlib axis
"""
ax = ax or plt.gca()
image = self.get_image(img_path)
ax.imshow(
image,
extent=[self.x_0, self.x_1, self.y_0, self.y_1],
origin="lower",
zorder=-100,
**kwargs,
)
if set_ax:
ax.set(
xlim=[self.x_0, self.x_1],
ylim=[self.y_0, self.y_1],
xlabel="cm",
ylabel="cm",
xticks=[self.x_0, self.x_1],
yticks=[self.y_0, self.y_1],
)
clean_axes(ax.figure)
ax.axis("equal")
def plot_n_units_per_channel(rname, units, rsites, TARGETS):
"""
Plot the number of units on each channel from a single recording, highlighting
channels in specific regoins
"""
logger.info(f"Plotting n units per channel for {rname}")
f, axes = plt.subplots(figsize=(12, 12), ncols=2, sharey=True)
f.suptitle(rname)
f._save_name = f"activity_units_per_channel"
# draw probe
plot_probe_electrodes(rsites, axes[0], TARGETS)
# draw barplot of # units per channel
counts = units.groupby("site_id").count()["name"]
_colors = [
rsites.loc[rsites.site_id == n]["color"].iloc[0] for n in counts.index
]
_regions = [
rsites.loc[rsites.site_id == n]["brain_region"].iloc[0]
for n in counts.index
]
colors = [
c if r in TARGETS else ("k" if r in ("unknown", "OUT") else blue_grey)
for c, r in zip(_colors, _regions)
]
probe_coords = [
rsites.loc[rsites.site_id == n]["probe_coordinates"].iloc[0]
for n in counts.index
]
axes[1].scatter(
counts.values + np.random.normal(0, 0.02, size=len(counts.values)),
probe_coords,
color=colors,
s=100,
lw=1,
ec="k",
)
for x, y in zip(counts.values, probe_coords):
axes[1].plot([0, x], [y, y], color=[0.2, 0.2, 0.2], lw=2, zorder=-1)
# cleanup and save
axes[0].set(
ylabel="Probe position (um)",
xticks=[],
xlim=[0.5, 1.5],
ylim=[0, 8000],
)
axes[1].set(xlabel="# units per channel", ylim=[0, 8000])
clean_axes(f)
return f
def plot_unit(
mouse_id: str,
session_name: str,
tracking: pd.DataFrame,
bouts: pd.DataFrame,
out_bouts: pd.DataFrame,
in_bouts: pd.DataFrame,
bouts_stacked: pd.DataFrame,
units: pd.DataFrame,
unit_id: Union[int, str],
tone_onsets: np.ndarray,
WINDOW: int,
):
frames = np.arange(0, len(tracking.x), 60 * 60 * 5)
time = (frames / 60 / 60).astype(np.int32)
if isinstance(unit_id, int):
if unit_id not in units.unit_id.values:
raise ValueError(f"Unit id {unit_id} not in units list:\n{units}")
for i, unit in units.iterrows():
if isinstance(unit_id, int):
if unit.unit_id != unit_id:
continue
logger.info(
f'Showing activity summary for unit unit {i+1}/{len(units)} (id: {unit.unit_id} - in: "{unit.brain_region}")'
)
# get tracking data at each spike
tracking["firing_rate"] = unit.firing_rate
unit_tracking = data_utils.select_by_indices(tracking, unit.spikes)
unit_tracking["spikes"] = unit.spikes
unit_vmax_frate = np.percentile(unit.firing_rate, 98)
out_bouts_stacked = data_utils.get_bouts_tracking_stacked(
tracking, out_bouts)
in_bouts_stacked = data_utils.get_bouts_tracking_stacked(
tracking, in_bouts)
# crate figure
f = plt.figure(figsize=(24, 12))
axes = f.subplot_mosaic("""
ARBCDUV
ARBCEUV
FSGHIXX
FSGHLXX
MTNOPZY
MTNOQZY
""")
f.suptitle(session_name + f"unit {unit.unit_id} {unit.brain_region}")
f._save_name = f"unit_{unit.unit_id}_{unit.brain_region}".replace(
"\\", "_")
# plot spikes against tracking, speed and angular velocity
visuals.plot_heatmap_2d(unit_tracking,
"spikes",
axes["A"],
cmap="inferno",
vmax=None)
axes["B"].plot(tracking.speed, color=blue_grey, lw=2)
axes["B"].scatter(
unit.spikes,
unit_tracking.speed,
color=colors.speed,
s=5,
zorder=11,
)
axes["C"].plot(tracking.dmov_velocity, color=blue_grey, lw=2)
axes["C"].scatter(
unit.spikes,
unit_tracking.dmov_velocity,
color=colors.dmov_velocity,
s=5,
zorder=11,
)
# plot spikes heatmap
visuals.plot_heatmap_2d(
unit_tracking,
"firing_rate",
axes["R"],
cmap="inferno",
vmax=unit_vmax_frate,
)
# plot spikes raster around tone onsets
visuals.plot_raster(unit.spikes, tone_onsets, axes["D"], window=WINDOW)
visuals.plot_aligned(
tracking.firing_rate,
tone_onsets,
axes["E"],
"aft",
color=blue_grey,
lw=1,
alpha=0.85,
window=WINDOW,
)
# plot spike rasters at bouts onsets and offsets
visuals.plot_raster(unit.spikes,
bouts.start_frame,
axes["I"],
window=WINDOW)
visuals.plot_aligned(
tracking.firing_rate,
bouts.start_frame,
axes["L"],
"aft",
color=blue_grey,
lw=1,
alpha=0.85,
window=WINDOW,
)
visuals.plot_raster(unit.spikes,
bouts.end_frame,
axes["P"],
window=WINDOW)
visuals.plot_aligned(
tracking.firing_rate,
bouts.end_frame,
axes["Q"],
"pre",
color=blue_grey,
lw=1,
alpha=0.85,
window=WINDOW,
)
# plot firing rate binned by speed and angular velocity
visuals.plot_bin_x_by_y(
tracking,
"firing_rate",
"speed",
axes["U"],
colors=colors.speed,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
tracking,
x_key="speed",
y_key="firing_rate",
ax=axes["U"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
visuals.plot_bin_x_by_y(
tracking,
"firing_rate",
"dmov_velocity",
axes["V"],
colors=colors.dmov_velocity,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
tracking,
x_key="dmov_velocity",
y_key="firing_rate",
ax=axes["V"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
axes["H"].axvline(0, ls=":", lw=2, color=[0.2, 0.2, 0.2], zorder=101)
# plot probe electrodes in which there is the unit
visuals.plot_probe_electrodes(
db_tables.Unit.get_unit_sites(mouse_id, session_name,
unit["unit_id"],
unit['probe_configuration']),
axes["Z"],
annotate_every=1,
TARGETS=None,
x_shift=False,
s=100,
lw=2,
)
# plot firing rate based on movements
visuals.plot_avg_firing_rate_based_on_movement(tracking, unit,
axes["X"])
# plot heatmap of firing rate vs speed by ang vel heatmap (during bouts)
trk = dict(
speed=tracking.speed[tracking.walking == 1],
dmov_velocity=tracking.dmov_velocity[tracking.walking == 1],
firing_rate=tracking.firing_rate[tracking.walking == 1],
)
visuals.plot_heatmap_2d(
trk,
key="firing_rate",
ax=axes["Y"],
x_key="speed",
y_key="dmov_velocity",
vmax=None,
)
# --------------------------------- in bouts --------------------------------- #
# plot bouts 2d
visuals.plot_bouts_heatmap_2d(tracking,
in_bouts,
"firing_rate",
axes["F"],
vmax=unit_vmax_frate)
# plot firing rate binned by global coordinates
visuals.plot_bin_x_by_y(
in_bouts_stacked,
"firing_rate",
"global_coord",
axes["S"],
colors=colors.global_coord,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
in_bouts_stacked,
x_key="global_coord",
y_key="firing_rate",
ax=axes["S"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
# plot firing rate binned by speed and angular velocity
visuals.plot_bin_x_by_y(
in_bouts_stacked,
"firing_rate",
"speed",
axes["G"],
colors=colors.speed,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
in_bouts_stacked,
x_key="speed",
y_key="firing_rate",
ax=axes["G"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
visuals.plot_bin_x_by_y(
in_bouts_stacked,
"firing_rate",
"dmov_velocity",
axes["H"],
colors=colors.dmov_velocity,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
in_bouts_stacked,
x_key="dmov_velocity",
y_key="firing_rate",
ax=axes["H"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
axes["H"].axvline(0, ls=":", lw=2, color=[0.2, 0.2, 0.2], zorder=101)
# --------------------------------- out bouts -------------------------------- #
visuals.plot_bouts_heatmap_2d(tracking,
out_bouts,
"firing_rate",
axes["M"],
vmax=unit_vmax_frate)
# plot firing rate binned by global coordinates
visuals.plot_bin_x_by_y(
out_bouts_stacked,
"firing_rate",
"global_coord",
axes["T"],
colors=colors.global_coord,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
x_key="global_coord",
y_key="firing_rate",
ax=axes["T"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
# plot firing rate binned by speed and angular velocity
visuals.plot_bin_x_by_y(
out_bouts_stacked,
"firing_rate",
"speed",
axes["N"],
colors=colors.speed,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
x_key="speed",
y_key="firing_rate",
ax=axes["N"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
visuals.plot_bin_x_by_y(
out_bouts_stacked,
"firing_rate",
"dmov_velocity",
axes["O"],
colors=colors.dmov_velocity,
bins=10,
min_count=10,
s=50,
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
x_key="dmov_velocity",
y_key="firing_rate",
ax=axes["O"],
vmax=None,
zorder=-10,
alpha=0.5,
cmap="inferno",
linewidths=0,
gridsize=20,
)
axes["H"].axvline(0, ls=":", lw=2, color=[0.2, 0.2, 0.2], zorder=101)
# ----------------------------- cleanup and save ----------------------------- #
clean_axes(f)
set_figure_subplots_aspect(wspace=0.5, hspace=0.6, left=0.3)
move_figure(f, 50, 50)
f.tight_layout()
axes["A"].set(
xlabel="xpos (cm)",
ylabel="ypos (cm)",
title=
f"unit {unit.unit_id} {unit.brain_region} | {len(unit.spikes)} spikes",
)
axes["B"].set(
xticks=frames,
xticklabels=time,
xlabel="time (min)",
ylabel="speed (cm/s)",
)
axes["C"].set(
xticks=frames,
xticklabels=time,
xlabel="time (min)",
ylabel="ang vel (deg/s)",
)
axes["D"].set(title="tone onset")
axes["F"].set(title="firing rate", ylabel="IN BOUTS")
axes["G"].set(ylabel="firing rate", xlabel="speed")
axes["H"].set(ylabel="firing rate",
xlabel="angular velocity",
xlim=[-350, 350])
axes["I"].set(title="bout ONSET")
# axes['L'].set(ylabel='firing rate', xlabel='global coord', xticks=np.arange(0, 1.1, .25))
axes["M"].set(title="firing rate", ylabel="OUT BOUTS")
axes["N"].set(ylabel="firing rate", xlabel="speed")
axes["O"].set(ylabel="firing rate",
xlabel="angular velocity",
xlim=[-350, 350])
axes["P"].set(title="bouts OFFSET")
# axes['Q'].set(ylabel='firing rate', xlabel='angular vel')
axes["R"].set(title="firing rate")
axes["S"].set(ylabel="firing rate", xlabel="global coord")
axes["T"].set(ylabel="firing rate", xlabel="global coord")
axes["U"].set(ylabel="firing rate", xlabel="speed")
axes["V"].set(ylabel="firing rate",
xlabel="angular velocity",
xlim=[-350, 350])
axes["Z"].set(xticks=[])
axes["Y"].set(
title="bouts firing rate heatmap",
xlabel="speed (cm/s)",
ylabel="ang vel (deg/s)",
)
for ax in "ARFM":
axes[ax].axis("equal")
axes[ax].set(xlim=[-5, 45],
xticks=[0, 40],
ylim=[-5, 65],
yticks=[0, 60])
def plot_hairpin_tracking(
session_name: str,
tracking: pd.DataFrame,
downsampled_tracking: pd.DataFrame,
bouts: pd.DataFrame,
out_bouts: pd.DataFrame,
in_bouts: pd.DataFrame,
bouts_stacked: pd.DataFrame,
out_bouts_stacked: pd.DataFrame,
in_bouts_stacked: pd.DataFrame,
):
# crate figure
f = plt.figure(figsize=(24, 12))
axes = f.subplot_mosaic("""
ABCDERX
ABCDERX
FGPHHSY
FGPIITY
LMQNNUW
LMQOOVW
""")
f.suptitle(session_name)
f._save_name = "tracking_data_2d"
# draw tracking 2D
visuals.plot_tracking_xy(
downsampled_tracking,
ax=axes["A"],
plot=True,
color=[0.6, 0.6, 0.6],
alpha=0.5,
)
visuals.plot_bouts_2d(tracking, bouts, axes["A"], lw=2, zorder=100)
# draw tracking 1D
visuals.plot_tracking_linearized(downsampled_tracking,
ax=axes["B"],
plot=True,
color=[0.6, 0.6, 0.6])
visuals.plot_bouts_1d(tracking, bouts, axes["B"], lw=3, zorder=100)
# plot speed aligned to bouts starts and ends
visuals.plot_aligned(tracking.speed,
bouts.start_frame,
axes["C"],
"after",
alpha=0.5)
visuals.plot_aligned(tracking.speed,
bouts.end_frame,
axes["D"],
"pre",
alpha=0.5)
# plot histograms of bouts durations
axes["E"].hist(
out_bouts.duration,
color=colors.outbound,
label="out",
bins=15,
alpha=0.7,
ec=[0.2, 0.2, 0.2],
histtype="stepfilled",
lw=2,
)
axes["E"].hist(
in_bouts.duration,
color=colors.inbound,
label="in",
bins=15,
alpha=0.7,
ec=[0.2, 0.2, 0.2],
histtype="stepfilled",
lw=2,
)
# plot histograms of speed and angular velocity distribution
axes["X"].hist(tracking.speed, bins=50, color=colors.speed, density=True)
axes["X"].axvline(
db_tables.Movement.moving_threshold,
lw=2,
ls="--",
color=grey_dark,
zorder=100,
)
axes["Y"].hist(
tracking.angular_velocity,
bins=50,
color=colors.angular_velocity,
density=True,
)
axes["Y"].axvline(
db_tables.Movement.turning_threshold,
lw=2,
ls="--",
color=grey_dark,
zorder=100,
)
axes["Y"].axvline(
-db_tables.Movement.turning_threshold,
lw=2,
ls="--",
color=grey_dark,
zorder=100,
)
# plot speed vs angular velocity inn all data and during botus only
trk = dict(
speed=tracking.speed[tracking.moving == 1],
angular_velocity=tracking.angular_velocity[tracking.moving == 1],
)
visuals.plot_heatmap_2d(
trk,
key=None,
ax=axes["R"],
x_key="speed",
y_key="angular_velocity",
vmax=None,
)
visuals.plot_heatmap_2d(
bouts_stacked,
key=None,
ax=axes["W"],
x_key="speed",
y_key="angular_velocity",
vmax=None,
)
# draw speed and orientation heatmaps during bouts
visuals.plot_heatmap_2d(
in_bouts_stacked,
"speed",
ax=axes["F"],
alpha=1,
vmax=30,
cmap="inferno",
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
"speed",
ax=axes["L"],
alpha=1,
vmax=30,
cmap="inferno",
)
visuals.plot_heatmap_2d(
in_bouts_stacked,
"orientation",
ax=axes["G"],
alpha=1,
vmin=0,
vmax=360,
edgecolors=grey_darker,
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
"orientation",
ax=axes["M"],
alpha=1,
vmin=0,
vmax=360,
edgecolors=grey_darker,
)
visuals.plot_heatmap_2d(
in_bouts_stacked,
"angular_velocity",
ax=axes["P"],
alpha=1,
vmin=-45,
vmax=45,
edgecolors=grey_darker,
)
visuals.plot_heatmap_2d(
out_bouts_stacked,
"angular_velocity",
ax=axes["Q"],
alpha=1,
vmin=-45,
vmax=45,
edgecolors=grey_darker,
)
# plot speeds binned by global coords for in/out bouts
nbins = 25
clrs = make_palette(grey_light, grey_dark, nbins - 1)
clrs2 = make_palette(amber_light, amber_darker, nbins - 1)
visuals.plot_bin_x_by_y(
in_bouts_stacked,
"speed",
"global_coord",
axes["H"],
bins=np.linspace(0, 1, nbins),
colors=clrs,
)
visuals.plot_bin_x_by_y(
in_bouts_stacked,
"angular_velocity",
"global_coord",
axes["I"],
bins=np.linspace(0, 1, nbins),
colors=clrs2,
)
visuals.plot_bin_x_by_y(
out_bouts_stacked,
"speed",
"global_coord",
axes["N"],
bins=np.linspace(0, 1, nbins),
colors=clrs,
)
visuals.plot_bin_x_by_y(
out_bouts_stacked,
"angular_velocity",
"global_coord",
axes["O"],
bins=np.linspace(0, 1, nbins),
colors=clrs2,
)
# plot histograms with speed and angular velocity
axes["S"].hist(in_bouts_stacked.speed, bins=20, color=colors.speed)
axes["T"].hist(
in_bouts_stacked.angular_velocity,
bins=20,
color=colors.angular_velocity,
)
axes["U"].hist(out_bouts_stacked.speed, bins=20, color=colors.speed)
axes["V"].hist(
out_bouts_stacked.angular_velocity,
bins=20,
color=colors.angular_velocity,
)
# ---------------------------------- cleanup --------------------------------- #
for ax in "IO":
axes[ax].axhline(0, lw=1, color=[0.6, 0.6, 0.6], zorder=-1)
# plot bouts centered
# visuals.plot_bouts_x_by_y(tracking, in_bouts, axes['S'], 'speed', 'global_coord')
# cleanup and save
clean_axes(f)
move_figure(f, 50, 50)
f.tight_layout()
axes["A"].set(
xlabel="xpos (cm)",
ylabel="ypos (cm)",
title=f"{len(in_bouts)} IN - {len(out_bouts)} OUT",
)
axes["B"].set(ylabel="time in exp", xlabel="arena position")
axes["C"].set(
ylabel="speed (cm/s)",
xticks=[0, 60, 120],
xticklabels=[-1, 0, 1],
xlabel="time (s)",
title="bout onset",
)
axes["D"].set(
ylabel="speed (cm/s)",
xticks=[0, 60, 120],
xticklabels=[-1, 0, 1],
xlabel="time (s)",
title="bout offset",
)
axes["E"].set(ylabel="counts",
xlabel="duration (s)",
title="Bouts duration")
axes["F"].set(ylabel="IN", xticks=[], yticks=[], title="speed")
axes["G"].set(xticks=[], yticks=[], title="orientation")
axes["H"].set(ylabel="speed (cm/s)", xticks=[])
axes["I"].set(ylabel="ang vel (deg/s)", xticks=[])
axes["L"].set(ylabel="OUT", xticks=[], yticks=[], xlabel="speed")
axes["M"].set(xticks=[], yticks=[], xlabel="orientation")
axes["N"].set(ylabel="speed (cm/s)", xticks=[])
axes["O"].set(ylabel="ang vel (deg/s)", xticks=np.linspace(0, 1, 11))
axes["P"].set(xticks=[], yticks=[], title="angular velocity")
axes["Q"].set(xticks=[], yticks=[], xlabel="ang vel (deg/s)")
axes["R"].set(xlabel="speed (cm/s)",
ylabel="ang vel (deg/s)",
title="moving")
axes["S"].set(xlabel="speed (cm/s)", title="in bouts")
axes["T"].set(xlabel="angular velocity (deg/s)")
axes["U"].set(xlabel="speed (cm/s)", title="out bouts")
axes["V"].set(xlabel="angular velocity (deg/s)")
axes["X"].set(xlabel="speed (cm/s)", title="all tracking")
axes["Y"].set(xlabel="angular velocity (deg/s)")
axes["W"].set(xlabel="speed (cm/s)",
ylabel="ang vel (deg/s)",
title="bouts heatmap")
for ax in "AFGLMPQ":
axes[ax].axis("equal")
axes[ax].set(xlim=[-5, 45],
ylim=[-5, 65],
xticks=[0, 40],
yticks=[0, 60])
zorder=100,
alpha=0.5)
# plot against speed
axes_dict["C"].plot(time, speed, color=blue_grey)
axes_dict["C"].scatter(
time[spike_times],
speed[spike_times],
color=color,
zorder=100,
alpha=0.5,
)
# plot against speed
axes_dict["D"].plot(time, avel, color=blue_grey)
axes_dict["D"].scatter(
time[spike_times],
avel[spike_times],
color=color,
zorder=100,
alpha=0.5,
)
# cleanup and save
axes_dict["A"].set(xlabel="xpos (cm)", ylabel="ypos (cm)")
axes_dict["C"].set(xlabel="time (frames)", ylabel="speed (cm/s)")
clean_axes(f)
plt.show()
# break
def plot_unit_firing_rate(unit: pd.Series, end: int = 60):
"""
For a single unit plot the firing rate at each moment for different firing rate windows
"""
if unit.empty:
raise ValueError(
"An empty pandas series was passed, perhaps the unit ID was invalid."
)
# get the unit data
name = unit["name"]
data = pd.DataFrame((Unit * Unit.Spikes * FiringRate
& f'name="{name}"'
& f"unit_id={unit.unit_id}").fetch())
frate_windows = data.firing_rate_std.values
n_frate_windows = len(frate_windows)
logger.info(
f"Found {n_frate_windows} firing rate windows to plot for unit {unit.unit_id}: {frate_windows}"
)
# create figure
f, axes = plt.subplots(nrows=2, sharex=True, figsize=(16, 9))
f.suptitle(f"Unit {unit.unit_id} firing rate")
f._save_title = f"unit_{unit.unit_id}_firing_rate"
# plot spikes
spikes = unit.spikes[unit.spikes < end * 60]
axes[0].hist(spikes, bins=end * 10, color=[0.5, 0.5, 0.5], density=True)
axes[0].scatter(
spikes,
np.random.uniform(-0.05, -0.001, size=len(spikes)),
color=black,
s=25,
zorder=1,
label="spike times",
)
palette = make_palette(orange_dark, blue_light, n_frate_windows)
for frate, color in zip(frate_windows, palette):
if frate != 100:
continue
frate_data = (
data.loc[data.firing_rate_std == frate].iloc[0].firing_rate[:end *
60])
axes[1].plot(frate_data,
lw=2,
color=color,
label=f"kernel std: {frate} ms")
# break
# cleanup
clean_axes(f)
time = np.arange(0, (end + 1) * 60, 60 * 2)
axes[1].legend()
axes[1].set(
xticks=time,
xticklabels=(time / 60).astype(np.int32),
xlabel="time (s)",
ylabel="firing rate",
)
