import matplotlib.pyplot as plt
import h5py
import numpy as np
import imageio
from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvas
[docs]def imaging_4plot(filepath, title, datetime, doppler, rng, snr, az, el):
"""
Parameters
----------
filepath
title
datetime
doppler
rng
snr
az
el
Returns
-------
Todo
* this will break when i move level 2 data to hdf5 format.
"""
az *= -1
doppler = np.where(doppler >= 50, (doppler - 100) * 10 * 3, doppler * 10 * 3)
# Method: remove the arc curvature of the earth.
pre_alt = np.sqrt(6378**2+(rng*0.75-200)**2 - 2*6378*(rng*0.75-200)*np.cos(np.deg2rad(90 + np.abs(el))))
gamma = np.arccos(((rng*0.75-200)**2 - (6378**2) - (pre_alt**2))/(-2*6378*pre_alt))
el = np.abs(el) - np.abs(np.rad2deg(gamma))
el = np.where(el > 12, np.nan, el)
alt = -6378+np.sqrt(6378**2+(rng*0.75-200)**2 - 2*6378*(rng*0.75-200)*np.cos(np.deg2rad(90 + np.abs(el))))
# North-South and East-West determination
rng = np.where(az >= 180, (rng * 0.75 - 200) * -1, rng * 0.75 - 200)
r = rng * np.cos(np.deg2rad(np.abs(el)))
horz = np.abs(r) * np.sin(np.deg2rad(az))
r *= np.cos(np.deg2rad(az))
# Clutter floor filtering
r = np.where(alt < 60, np.nan, r) # 60 or 85
horz = np.where(alt < 60, np.nan, horz) # 60 or 85
alt = np.where(alt < 60, np.nan, alt) # 60 or 85
# Setup plotting area.
plt.figure(x + 1, figsize=[12, 13])
plt.rcParams.update({'font.size': 20})
plt.suptitle(title + ' ' + datetime)
# Top down view with Doppler.
plt.subplot(221)
plt.grid()
plt.ylabel('South-North Distance [km]')
plt.scatter(horz, r, c=doppler, cmap='jet_r', vmin=-1000, vmax=1000, alpha=0.5)
plt.xlim((-400, 400))
plt.ylim((0, 1000))
# Side view with Doppler.
plt.subplot(222)
plt.grid()
plt.plot(np.ones(len(rng)) * 130, np.linspace(0, 1000, len(r)), '--k', zorder=1)
plt.plot(np.ones(len(rng)) * 80, np.linspace(0, 1000, len(r)), '--k', zorder=1)
plt.scatter(pm * alt, r, c=doppler, cmap='jet_r', vmin=-1000, vmax=1000, zorder=2, alpha=0.5)
plt.colorbar(label='Doppler Velocity [m/s]')
plt.xlim((0, 200))
plt.ylim((0, 1000))
# Top down view with SNR.
plt.subplot(223)
plt.grid()
plt.ylabel('South-North Distance [km]')
plt.xlabel('West-East Distance [km]')
plt.scatter(horz, r, c=snr, cmap='plasma_r', vmin=0, vmax=20, alpha=0.5)
plt.xlim((-400, 400))
plt.ylim((0, 1000))
# Side view with SNR.
plt.subplot(224)
plt.grid()
plt.xlabel('Corrected Altitude [km]')
plt.plot(np.ones(len(rng)) * 130, np.linspace(0, 1000, len(r)), '--k', zorder=1)
plt.plot(np.ones(len(rng)) * 80, np.linspace(0, 1000, len(r)), '--k', zorder=1)
plt.scatter(pm * alt, r, c=snr, cmap='plasma_r', vmin=0, vmax=20, zorder=2, alpha=0.5)
plt.colorbar(label='Signal-to-Noise [dB]')
plt.xlim((0, 200))
plt.ylim((0, 1000))
plt.tight_layout(rect=[0, 0.03, 1, 0.95])
plt.savefig(filepath + str(title + datetime).replace(' ', '_').replace(':', '') + '.png')
plt.close()
return None
[docs]def quick_look(config, time):
"""
Creates a standard Quick Look plot of level 1 data for the specified time frame.
Parameters
----------
config : Class Object
Config class instantiation which contains plotting settings.
time : Class Object
Time class instantiation for start, stop, step deceleration.
Returns
-------
None
Notes
-----
* Typically a Quick Look plot should be one day of data with a step size equal to the incoherent averages
time length used to generate the level 1 data used.
Todo
* The plt.colorbar() is currently not working.
"""
plt.figure(1, figsize=[20, 10])
plt.rcParams.update({'font.size': 22})
temp_hour = [-1, -1, -1, -1]
for t in range(int(time.start_epoch), int(time.stop_epoch), int(time.step_epoch)):
now = time.get_date(t)
if [int(now.year), int(now.month), int(now.day), int(now.hour)] != temp_hour:
try:
filename = h5py.File(f'{config.plotting_source}{config.radar_name}_{config.processing_method}_'
f'{config.tx_name}_{config.rx_name}_'
f'{int(config.snr_cutoff):02d}dB_{config.incoherent_averages:02d}00ms_'
f'{int(now.year):04d}_'
f'{int(now.month):02d}_'
f'{int(now.day):02d}_'
f'{int(now.hour):02d}.h5', 'r')
except:
continue
temp_hour = [int(now.year), int(now.month), int(now.day), int(now.hour)]
try:
moment = f'data/{int(now.hour):02d}{int(now.minute):02d}{int(now.second * 1000):05d}'
if bool(filename[f'{moment}/data_flag']):
tau = int((now.hour * 60 * 60 + now.minute * 60 + now.second / 1000) / 3600)
snr_db = np.abs(filename[f'{moment}/snr_db'][:])
doppler_shift = filename[f'{moment}/doppler_shift'][:]
rf_distance = np.abs(filename[f'{moment}/rf_distance'][:])
plt.subplot(212)
plt.scatter(np.ones(len(rf_distance)) * tau, rf_distance, c=doppler_shift * 3.03,
vmin=-900.0, vmax=900.0, s=3, cmap='jet_r')
plt.colorbar(label='Doppler (m/s)')
plt.subplot(211)
plt.scatter(np.ones(len(rf_distance)) * tau, rf_distance, c=snr_db, vmin=0.0,
vmax=100.0, s=3, cmap='plasma_r')
plt.colorbar(label='SNR (dB)')
except:
continue
plt.subplot(211)
plt.title(f'{int(time.start_human.year):04d}-{int(time.start_human.month):02d}-{int(time.start_human.day):02d}'
f' {config.radar_name} Quick Look Plot')
plt.ylabel('RF Distance (km)')
plt.ylim(0, 2500)
plt.xlim(0.0, 24.0)
plt.grid()
plt.subplot(212)
plt.xlabel('Time (hours)')
plt.ylabel('RF Distance (km)')
plt.ylim(0, 2500)
plt.xlim(0.0, 24.0)
plt.grid()
plt.savefig(f'{config.plotting_destination}quicklook_{config.radar_name}_'
f'{int(time.start_human.year):04d}-'
f'{int(time.start_human.month):02d}-'
f'{int(time.start_human.day):02d}.png')
plt.close(1)
return None
[docs]def range_doppler_snr(config, time, spacing):
"""
Creates a standard range-Doppler SNR plot of level 1 data for the specified time frame.
Parameters
----------
config : Class Object
Config class instantiation which contains plotting settings.
time : Class Object
Time class instantiation for start, stop, step deceleration.
spacing : int
The amount of time in seconds to plot in one image.
Returns
-------
None
Notes
-----
* Typically a Quick Look plot should be one day of data with a step size equal to the incoherent averages
time length used to generate the level 1 data used.
"""
temp_hour = [-1, -1, -1, -1]
spacing_counter = 0
data_flag = False
with imageio.get_writer(f'{config.plotting_destination}{config.radar_config}_{config.experiment_name}_'
f'range_doppler_snr_{spacing}sec_movie_'
f'{int(time.start_human.year):04d}_'
f'{int(time.start_human.month):02d}_'
f'{int(time.start_human.day):02d}_'
f'{int(time.start_human.hour)}'
f'.mp4', fps=10, mode='I') as writer:
for t in range(int(time.start_epoch), int(time.stop_epoch), int(time.step_epoch)):
now = time.get_date(t)
if [int(now.year), int(now.month), int(now.day), int(now.hour)] != temp_hour:
print(f'{config.plotting_source}{config.radar_config}_{config.experiment_name}_'
f'{int(config.snr_cutoff_db):02d}dB_{config.incoherent_averages:02d}00ms_'
f'{int(now.year):04d}_'
f'{int(now.month):02d}_'
f'{int(now.day):02d}_'
f'{int(now.hour):02d}_'
f'{config.tx_site_name}_{config.rx_site_name}.h5')
try:
filename = h5py.File(f'{config.plotting_source}{config.radar_config}_{config.experiment_name}_'
f'{int(config.snr_cutoff_db):02d}dB_{config.incoherent_averages:02d}00ms_'
f'{int(now.year):04d}_'
f'{int(now.month):02d}_'
f'{int(now.day):02d}_'
f'{int(now.hour):02d}_'
f'{config.tx_site_name}_{config.rx_site_name}.h5', 'r')
except:
print('ERROR: File skipped or does not exist')
continue
temp_hour = [int(now.year), int(now.month), int(now.day), int(now.hour)]
spacing_counter += 1
plt.figure(1)
# Save the image if over spacing
if spacing_counter > spacing:
spacing_counter = 1
if data_flag:
print('\tsaving image:', save_name)
plt.savefig(save_name + '.pdf')
fig = plt.figure(1)
canvas = FigureCanvas(fig)
canvas.draw()
writer.append_data(np.asarray(canvas.buffer_rgba()))
# plt.savefig(save_name + '.png')
plt.close(1)
plt.figure(1)
data_flag = False
# Start a new image
if spacing_counter == 1:
plt.scatter(0, -1, c=0.0, vmin=0.0, vmax=30.0, s=3, cmap='plasma_r')
plt.title(f'ICEBEAR-3D Range-Doppler-SNR Plot\n'
f'{int(now.year):04d}-'
f'{int(now.month):02d}-'
f'{int(now.day):02d} '
f'{int(now.hour):02d}:'
f'{int(now.minute):02d}:'
f'{int(now.second):02d}')
cb = plt.colorbar(label='SNR (dB)')
cb.ax.plot([-500, 500], [config.snr_cutoff_db, config.snr_cutoff_db], 'k')
plt.xlabel('Doppler (Hz)')
plt.ylabel('Total RF Distance (km)')
plt.ylim(0, config.number_ranges * config.range_resolution)
plt.xlim(-500, 500)
plt.xticks(np.arange(-500, 500 + 100, 100))
plt.yticks(np.arange(0, int(config.number_ranges * config.range_resolution + 500), 500))
plt.grid(linestyle=':')
props = dict(boxstyle='square', facecolor='wheat', alpha=0.5)
plt.text(-450, 150, f'{config.snr_cutoff_db} dB SNR Cutoff', bbox=props)
plt.text(250, 150, f'{spacing} s Interval', bbox=props)
save_name = f'{config.plotting_destination}{config.radar_config}_{config.experiment_name}_'\
f'range_doppler_snr_{spacing}sec_'\
f'{int(now.year):04d}_'\
f'{int(now.month):02d}_'\
f'{int(now.day):02d}_'\
f'{int(now.hour):02d}_'\
f'{int(now.minute):02d}_'\
f'{int(now.second):02d}'
# Add to the image if under spacing
if spacing_counter <= spacing:
try:
moment = f'data/{int(now.hour):02d}{int(now.minute):02d}{int(now.second * 1000):05d}'
if bool(filename[f'{moment}/data_flag']):
dop = filename[f'{moment}/doppler_shift'][:]
rng = np.abs(filename[f'{moment}/rf_distance'][:])
snr = np.abs(filename[f'{moment}/snr_db'][:])
plt.scatter(dop, rng, c=snr, vmin=0.0, vmax=30.0, s=3, cmap='plasma_r')
data_flag = True
except:
continue
writer.close()
return None