def test_grid_bounds():
"""Can we compute grid bounds correctly"""
lons = np.arange(-100, -80, 0.1)
lats = np.arange(29, 51, 0.2)
(x0, y0, x1, y1) = util.grid_bounds(lons, lats, [-96, 32, -89, 40])
assert x0 == 41
assert x1 == 111
assert y0 == 16
assert y1 == 56
(lons, lats) = np.meshgrid(lons, lats)
(x0, y0, x1, y1) = util.grid_bounds(lons, lats, [-96, 32, -89, 40])
assert x0 == 40
assert x1 == 110
assert y0 == 15
assert y1 == 55
def plotter(fdict):
""" Go """
ctx = util.get_autoplot_context(fdict, get_description())
ptype = ctx["ptype"]
sdate = ctx["sdate"]
edate = ctx["edate"]
src = ctx["src"]
opt = ctx["opt"]
usdm = ctx["usdm"]
if sdate.year != edate.year:
raise NoDataFound("Sorry, do not support multi-year plots yet!")
days = (edate - sdate).days
sector = ctx["sector"]
x0 = 0
x1 = -1
y0 = 0
y1 = -1
state = None
if len(sector) == 2:
state = sector
sector = "state"
title = compute_title(src, sdate, edate)
if src == "mrms":
ncfn = iemre.get_daily_mrms_ncname(sdate.year)
clncfn = iemre.get_dailyc_mrms_ncname()
ncvar = "p01d"
source = "MRMS Q3"
subtitle = "NOAA MRMS Project, GaugeCorr and RadarOnly"
elif src == "iemre":
ncfn = iemre.get_daily_ncname(sdate.year)
clncfn = iemre.get_dailyc_ncname()
ncvar = "p01d_12z"
source = "IEM Reanalysis"
subtitle = "IEM Reanalysis is derived from various NOAA datasets"
else:
ncfn = "/mesonet/data/prism/%s_daily.nc" % (sdate.year, )
clncfn = "/mesonet/data/prism/prism_dailyc.nc"
ncvar = "ppt"
source = "OSU PRISM"
subtitle = ("PRISM Climate Group, Oregon State Univ., "
"http://prism.oregonstate.edu, created 4 Feb 2004.")
mp = MapPlot(
sector=sector,
state=state,
axisbg="white",
nocaption=True,
title="%s:: %s Precip %s" % (source, title, PDICT3[opt]),
subtitle="Data from %s" % (subtitle, ),
titlefontsize=14,
)
idx0 = iemre.daily_offset(sdate)
idx1 = iemre.daily_offset(edate) + 1
if not os.path.isfile(ncfn):
raise NoDataFound("No data for that year, sorry.")
with util.ncopen(ncfn) as nc:
if state is not None:
x0, y0, x1, y1 = util.grid_bounds(
nc.variables["lon"][:],
nc.variables["lat"][:],
state_bounds[state],
)
elif sector in SECTORS:
bnds = SECTORS[sector]
x0, y0, x1, y1 = util.grid_bounds(
nc.variables["lon"][:],
nc.variables["lat"][:],
[bnds[0], bnds[2], bnds[1], bnds[3]],
)
lats = nc.variables["lat"][y0:y1]
lons = nc.variables["lon"][x0:x1]
if sdate == edate:
p01d = mm2inch(nc.variables[ncvar][idx0, y0:y1, x0:x1])
elif (idx1 - idx0) < 32:
p01d = mm2inch(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0))
else:
# Too much data can overwhelm this app, need to chunk it
for i in range(idx0, idx1, 10):
i2 = min([i + 10, idx1])
if idx0 == i:
p01d = mm2inch(
np.sum(nc.variables[ncvar][i:i2, y0:y1, x0:x1], 0))
else:
p01d += mm2inch(
np.sum(nc.variables[ncvar][i:i2, y0:y1, x0:x1], 0))
if np.ma.is_masked(np.max(p01d)):
raise NoDataFound("Data Unavailable")
plot_units = "inches"
cmap = get_cmap(ctx["cmap"])
cmap.set_bad("white")
if opt == "dep":
# Do departure work now
with util.ncopen(clncfn) as nc:
climo = mm2inch(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0))
p01d = p01d - climo
[maxv] = np.percentile(np.abs(p01d), [99])
clevs = np.around(np.linspace(0 - maxv, maxv, 11), decimals=2)
elif opt == "per":
with util.ncopen(clncfn) as nc:
climo = mm2inch(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0))
p01d = p01d / climo * 100.0
cmap.set_under("white")
cmap.set_over("black")
clevs = [1, 10, 25, 50, 75, 100, 125, 150, 200, 300, 500]
plot_units = "percent"
else:
p01d = np.where(p01d < 0.001, np.nan, p01d)
cmap.set_under("white")
clevs = [0.01, 0.1, 0.3, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 5, 6, 8, 10]
if days > 6:
clevs = [0.01, 0.3, 0.5, 1, 1.5, 2, 3, 4, 5, 6, 7, 8, 10, 15, 20]
if days > 29:
clevs = [0.01, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25, 30, 35]
if days > 90:
clevs = [0.01, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25, 30, 35, 40]
x2d, y2d = np.meshgrid(lons, lats)
if ptype == "c":
mp.contourf(x2d,
y2d,
p01d,
clevs,
cmap=cmap,
units=plot_units,
iline=False)
else:
res = mp.pcolormesh(x2d, y2d, p01d, clevs, cmap=cmap, units=plot_units)
res.set_rasterized(True)
if sector != "midwest":
mp.drawcounties()
mp.drawcities()
if usdm == "yes":
mp.draw_usdm(edate, filled=False, hatched=True)
return mp.fig
def plotter(fdict):
""" Go """
ctx = util.get_autoplot_context(fdict, get_description())
ptype = ctx['ptype']
sdate = ctx['sdate']
edate = ctx['edate']
src = ctx['src']
opt = ctx['opt']
usdm = ctx['usdm']
if sdate.year != edate.year:
raise NoDataFound('Sorry, do not support multi-year plots yet!')
days = (edate - sdate).days
sector = ctx['sector']
if sdate == edate:
title = sdate.strftime("%-d %B %Y")
else:
title = "%s to %s (inclusive)" % (sdate.strftime("%-d %b"),
edate.strftime("%-d %b %Y"))
x0 = 0
x1 = -1
y0 = 0
y1 = -1
state = None
if len(sector) == 2:
state = sector
sector = 'state'
if src == 'mrms':
ncfn = iemre.get_daily_mrms_ncname(sdate.year)
clncfn = iemre.get_dailyc_mrms_ncname()
ncvar = 'p01d'
source = 'MRMS Q3'
subtitle = 'NOAA MRMS Project, GaugeCorr and RadarOnly'
elif src == 'iemre':
ncfn = iemre.get_daily_ncname(sdate.year)
clncfn = iemre.get_dailyc_ncname()
ncvar = 'p01d_12z'
source = 'IEM Reanalysis'
subtitle = 'IEM Reanalysis is derived from various NOAA datasets'
else:
ncfn = "/mesonet/data/prism/%s_daily.nc" % (sdate.year, )
clncfn = "/mesonet/data/prism/prism_dailyc.nc"
ncvar = 'ppt'
source = 'OSU PRISM'
subtitle = ('PRISM Climate Group, Oregon State Univ., '
'http://prism.oregonstate.edu, created 4 Feb 2004.')
mp = MapPlot(sector=sector,
state=state,
axisbg='white',
nocaption=True,
title='%s:: %s Precip %s' % (source, title, PDICT3[opt]),
subtitle='Data from %s' % (subtitle, ),
titlefontsize=14)
idx0 = iemre.daily_offset(sdate)
idx1 = iemre.daily_offset(edate) + 1
if not os.path.isfile(ncfn):
raise NoDataFound("No data for that year, sorry.")
with util.ncopen(ncfn) as nc:
if state is not None:
x0, y0, x1, y1 = util.grid_bounds(nc.variables['lon'][:],
nc.variables['lat'][:],
state_bounds[state])
elif sector in SECTORS:
bnds = SECTORS[sector]
x0, y0, x1, y1 = util.grid_bounds(
nc.variables['lon'][:], nc.variables['lat'][:],
[bnds[0], bnds[2], bnds[1], bnds[3]])
lats = nc.variables['lat'][y0:y1]
lons = nc.variables['lon'][x0:x1]
if sdate == edate:
p01d = distance(nc.variables[ncvar][idx0, y0:y1, x0:x1],
'MM').value('IN')
elif (idx1 - idx0) < 32:
p01d = distance(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0),
'MM').value('IN')
else:
# Too much data can overwhelm this app, need to chunk it
for i in range(idx0, idx1, 10):
i2 = min([i + 10, idx1])
if idx0 == i:
p01d = distance(
np.sum(nc.variables[ncvar][i:i2, y0:y1, x0:x1], 0),
'MM').value('IN')
else:
p01d += distance(
np.sum(nc.variables[ncvar][i:i2, y0:y1, x0:x1], 0),
'MM').value('IN')
if np.ma.is_masked(np.max(p01d)):
raise NoDataFound("Data Unavailable")
units = 'inches'
cmap = plt.get_cmap(ctx['cmap'])
cmap.set_bad('white')
if opt == 'dep':
# Do departure work now
with util.ncopen(clncfn) as nc:
climo = distance(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0),
'MM').value('IN')
p01d = p01d - climo
[maxv] = np.percentile(np.abs(p01d), [
99,
])
clevs = np.around(np.linspace(0 - maxv, maxv, 11), decimals=2)
elif opt == 'per':
with util.ncopen(clncfn) as nc:
climo = distance(
np.sum(nc.variables[ncvar][idx0:idx1, y0:y1, x0:x1], 0),
'MM').value('IN')
p01d = p01d / climo * 100.
cmap.set_under('white')
cmap.set_over('black')
clevs = [1, 10, 25, 50, 75, 100, 125, 150, 200, 300, 500]
units = 'percent'
else:
p01d = np.where(p01d < 0.001, np.nan, p01d)
cmap.set_under('white')
clevs = [0.01, 0.1, 0.3, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 5, 6, 8, 10]
if days > 6:
clevs = [0.01, 0.3, 0.5, 1, 1.5, 2, 3, 4, 5, 6, 7, 8, 10, 15, 20]
if days > 29:
clevs = [0.01, 0.5, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25, 30, 35]
if days > 90:
clevs = [0.01, 1, 2, 3, 4, 5, 6, 8, 10, 15, 20, 25, 30, 35, 40]
x2d, y2d = np.meshgrid(lons, lats)
if ptype == 'c':
mp.contourf(x2d, y2d, p01d, clevs, cmap=cmap, units=units, iline=False)
else:
res = mp.pcolormesh(x2d, y2d, p01d, clevs, cmap=cmap, units=units)
res.set_rasterized(True)
if sector != 'midwest':
mp.drawcounties()
mp.drawcities()
if usdm == 'yes':
mp.draw_usdm(edate, filled=False, hatched=True)
return mp.fig