def plot_geo(geo, coastline, symbol, legend, title, code):
with NamedTemporaryFile(delete=False, suffix=f"__WT_{code}.pdf") as pdf:
pdf_obj = mv.pdf_output(output_name=pdf.name.replace(".pdf", ""))
mv.setoutput(pdf_obj)
mv.plot(coastline, symbol, legend, title, geo)
os.chmod(pdf.name, 0o444)
return {"pdf": pdf.name}
def plot_obs_freq(predictor_matrix, code):
coastline = mv.mcoast(map_coastline_thickness=2,
map_boundaries="on",
map_coastline_colour="chestnut")
symbol = mv.msymb(
legend="on",
symbol_type="marker",
symbol_table_mode="on",
symbol_outline="on",
symbol_min_table=[1, 2, 5, 10, 15, 20, 25, 30],
symbol_max_table=[2, 5, 10, 15, 20, 25, 30, 100000],
symbol_colour_table=[
"RGB(0.7020,0.7020,0.7020)",
"RGB(0.4039,0.4039,0.4039)",
"blue",
"RGB(0.4980,1.0000,0.0000)",
"RGB(1.0000,0.8549,0.0000)",
"orange",
"red",
"magenta",
],
symbol_marker_table=15,
symbol_height_table=0.3,
)
legend = mv.mlegend(
legend_text_font="arial",
legend_text_font_size=0.35,
legend_entry_plot_direction="row",
legend_box_blanking="on",
legend_entry_text_width=50,
)
title = mv.mtext(
text_line_count=4,
text_line_1=
"OBS Frequency", # To sostitute with "FE" values when relevant.
text_line_2=f"WT Code = {code}",
text_line_4=" ",
text_font="arial",
text_font_size=0.4,
)
df = predictor_matrix[["LonOBS", "LatOBS", "OBS"]]
grouped_df = df.groupby(["LatOBS", "LonOBS"], as_index=False).count()
geo = mv.create_geo(len(grouped_df), "xyv")
geo = mv.set_latitudes(geo, grouped_df["LatOBS"].to_numpy(dtype=np.float))
geo = mv.set_longitudes(geo, grouped_df["LonOBS"].to_numpy(dtype=np.float))
geo = mv.set_values(geo, grouped_df["OBS"].to_numpy(dtype=np.float))
with NamedTemporaryFile(delete=False, suffix=".pdf") as pdf:
pdf_obj = mv.pdf_output(output_name=pdf.name.replace(".pdf", ""))
mv.setoutput(pdf_obj)
mv.plot(coastline, symbol, legend, title, geo)
return pdf.name
def plot_std(predictor_matrix, code):
coastline = mv.mcoast(map_coastline_thickness=2,
map_boundaries="on",
map_coastline_colour="chestnut")
symbol = mv.msymb(
legend="on",
symbol_type="marker",
symbol_table_mode="on",
symbol_outline="on",
symbol_min_table=[0, 0.0001, 0.5, 1, 2, 5],
symbol_max_table=[0.0001, 0.5, 1, 2, 5, 1000],
symbol_colour_table=[
"RGB(0.7020,0.7020,0.7020)",
"RGB(0.2973,0.2973,0.9498)",
"RGB(0.1521,0.6558,0.5970)",
"RGB(1.0000,0.6902,0.0000)",
"red",
"RGB(1.0000,0.0000,1.0000)",
],
symbol_marker_table=15,
symbol_height_table=0.3,
)
legend = mv.mlegend(
legend_text_font="arial",
legend_text_font_size=0.35,
legend_entry_plot_direction="row",
legend_box_blanking="on",
legend_entry_text_width=50,
)
error = "FER" if "FER" in predictor_matrix.columns else "FE"
title = mv.mtext(
text_line_count=4,
text_line_1=f"{error} Standard Deviation",
text_line_2=f"WT Code = {code}",
text_line_4=" ",
text_font="arial",
text_font_size=0.4,
)
df = predictor_matrix[["LonOBS", "LatOBS", error]]
grouped_df = df.groupby(["LatOBS", "LonOBS"])[error].mean().reset_index()
geo = mv.create_geo(len(grouped_df), "xyv")
geo = mv.set_latitudes(geo, grouped_df["LatOBS"].to_numpy(dtype=np.float))
geo = mv.set_longitudes(geo, grouped_df["LonOBS"].to_numpy(dtype=np.float))
geo = mv.set_values(geo, grouped_df[error].to_numpy(dtype=np.float))
with NamedTemporaryFile(delete=False, suffix=".pdf") as pdf:
pdf_obj = mv.pdf_output(output_name=pdf.name.replace(".pdf", ""))
mv.setoutput(pdf_obj)
mv.plot(coastline, symbol, legend, title, geo)
return pdf.name
def test_plot_2_pages():
output_name = file_in_testdir('test_plot_2_pages')
png = mv.png_output(output_name=output_name)
degraded_field = mv.read(data=TEST_FIELDSET, grid=[4, 4])
page1 = mv.plot_page(top=2.2, bottom=52.2, right=100)
page2 = mv.plot_page(top=50)
dw = mv.plot_superpage(pages=[page1, page2])
mv.setoutput(png)
mv.plot(dw[0], degraded_field, dw[1], degraded_field + 50)
output_name_from_magics = output_name + '.1.png'
assert (os.path.isfile(output_name_from_magics))
os.remove(output_name_from_magics)
def plot_geo(geo, coastline, symbol, legend, title):
with NamedTemporaryFile(suffix=".png") as png, NamedTemporaryFile(
delete=False, suffix=".pdf") as pdf:
png_obj = mv.png_output(output_name=png.name.replace(".png", ""))
mv.setoutput(png_obj)
mv.plot(coastline, symbol, legend, title, geo)
pdf_obj = mv.pdf_output(output_name=pdf.name.replace(".pdf", ""))
mv.setoutput(pdf_obj)
mv.plot(coastline, symbol, legend, title, geo)
os.chmod(pdf.name, 0o444)
with open(png.name.replace(".png", ".1.png"), "rb") as img:
return {"preview": b64encode(img.read()).decode(), "pdf": pdf.name}
def build(self):
if not is_ipython_active():
return
img_size = 120
img, names = self._build(img_size)
# reset jupyter output settings
mv.setoutput("jupyter", **mv.plot.jupyter_args)
if len(img) > 0:
from IPython.display import HTML
return HTML(
self.build_gallery(names, img, row_height=f"{img_size}px"))
def test_plot_1():
output_name = file_in_testdir('test_plot_1')
mv.setoutput(mv.png_output(output_name=output_name))
grid_shade = {
'legend': True,
'contour': False,
'contour_highlight': True,
'contour_shade': True,
'contour_shade_technique': 'grid_shading',
'contour_shade_max_level_colour': 'red',
'contour_shade_min_level_colour': 'blue',
'contour_shade_colour_direction': 'clockwise',
}
degraded_field = mv.read(data=TEST_FIELDSET, grid=[4, 4])
mv.plot(degraded_field, mv.mcont(grid_shade))
output_name_from_magics = output_name + '.1.png'
assert (os.path.isfile(output_name_from_magics))
os.remove(output_name_from_magics)
def _build(self, img_size):
img = []
names = []
# img_size = 120
tmp_dir = os.path.join(os.getenv("METVIEW_TMPDIR", ""), "_maparea_")
Path(tmp_dir).mkdir(exist_ok=True)
for name in map_area_names():
f_name = os.path.join(tmp_dir, name + ".png")
if not os.path.exists(f_name):
view = mv.make_geoview(area=name, style="grey_1")
mv.setoutput(
mv.png_output(
output_name=f_name[:-4],
output_width=300,
output_name_first_page_number="off",
))
mv.plot(view)
if os.path.exists(f_name):
names.append(name)
img.append(self.to_base64(f_name))
return (img, names)
def _build(self, img_size):
img = []
names = []
# img_size = 120
tmp_dir = os.path.join(os.getenv("METVIEW_TMPDIR", ""), "_mapstyle_")
Path(tmp_dir).mkdir(exist_ok=True)
for name, d in map_styles().items():
f_name = os.path.join(tmp_dir, name + ".png")
if not os.path.exists(f_name):
view = mv.make_geoview(area=[30, -30, 75, 45], style=name)
view.update({"MAP_COASTLINE_RESOLUTION": "low"},
sub="coastlines")
mv.setoutput(
mv.png_output(
output_name=f_name[:-4],
output_width=300,
output_name_first_page_number="off",
))
mv.plot(view)
if os.path.exists(f_name):
names.append(name)
img.append(self.to_base64(f_name))
return (img, names)
map_coastline_sea_shade_colour="RGB(0.85,0.93,1)",
)
area_view = mv.geoview(
map_area_definition="corners",
area=[45.83, -13.87, 62.03, 8.92],
coastlines=land_shade,
)
# Simplest plot:
# NOTE that when plotting a 'raw' BUFR file, Magics will plot synop symbols as shown in
# https://software.ecmwf.int/wiki/display/METV/Data+Part+1 "Plotting BUFR Data"
obs = mv.read("../tests/obs_3day.bufr")
mv.setoutput(mv.png_output(output_width=1200, output_name="./obsplot1"))
mv.plot(area_view, obs)
# ALTERNATIVELY, add an Observations Plotting visual definition
obs_plotting = mv.mobs(
obs_temperature=False,
obs_cloud=False,
obs_low_cloud=False,
obs_dewpoint_colour="purple",
)
mv.setoutput(mv.png_output(output_width=1200, output_name="./obsplot2"))
mv.plot(area_view, obs, obs_plotting)
# ALTERNATIVELY, if we don't want to plot the whole observations, but instead want to
# read geopotential
z = mv.read("./z_for_UC-04.grib")
t_shade_c = mv.mcont(
legend=True,
contour_highlight=False,
contour_level_selection_type="interval",
contour_interval=10,
contour_shade=True,
contour_shade_max_level=60,
contour_shade_min_level=-60,
contour_shade_method="area_fill",
contour_shade_max_level_colour="red",
contour_shade_min_level_colour="blue",
contour_shade_colour_direction="clockwise",
)
z_isolines = mv.mcont(
legend=True,
contour_line_thickness=2,
contour_line_colour="black",
contour_highlight_colour="black",
contour_highlight_thickness=4,
contour_level_selection_type="interval",
contour_interval=5,
contour_legend_text="Geopotential",
)
mv.setoutput(mv.png_output(output_width=1000, output_name="./gribplot"))
mv.plot(t2, t_shade_c, z, z_isolines)
t2m_grib = mv.read("./t2m_grib.grib")
obs_3day = mv.read("./obs_3day.bufr")
t2m_gpt = mv.obsfilter(parameter="012004", output="geopoints", data=obs_3day)
diff = t2m_grib - t2m_gpt
diff_symb = mv.msymb(
legend=True,
symbol_type="marker",
symbol_table_mode="advanced",
)
mv.setoutput(mv.png_output(output_width=1000, output_name="./obsdiff1"))
mv.plot(area_view, diff, diff_symb)
# Extract geopoints that are hotter by 1 deg or more
# hotter = mv.filter(diff, diff >= 1)
hotter = diff.filter(diff >= 1)
# Extract geopoints that are colder by 1 deg or more
# colder = mv.filter(diff, diff <= -1)
colder = diff.filter(diff <= -1)
# Get geopoints that are within +/-1
# exact = mv.filter(diff, (diff > -1) * (diff < 1))
exact = diff.filter((diff > -1) * (diff < 1))
# Symbol visdefs for each classification
UC-05. The Analyst retrieves certain parameters, computes a derived
parameter and plots its values for a specific time.
--------------------------------------------------------------------------------
1. Analyst filters, the chosen parameters of a file from a given
source (i.e. MARS, files, observation databases)
--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
2. Analyst computes the desired derived parameter (i.e. wind velocity from zonal
and meridional components)
--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
3. Analyst plots the derived parameter values
--------------------------------------------------------------------------------
"""
import metview as mv
uv = mv.read('./uv.grib')
u = mv.read(data=uv, param='u')
v = mv.read(data=uv, param='v')
wind_speed = mv.sqrt(u * u + v * v)
mv.setoutput(mv.png_output(output_width=1000, output_name='./uvplot'))
mv.plot(wind_speed)
t2m_grib = mv.read('./t2m_grib.grib')
obs_3day = mv.read('./obs_3day.bufr')
t2m_gpt = mv.obsfilter(parameter='012004', output='geopoints', data=obs_3day)
diff = t2m_grib - t2m_gpt
diff_symb = mv.msymb(
legend=True,
symbol_type='marker',
symbol_table_mode='advanced',
)
mv.setoutput(mv.png_output(output_width=1000, output_name='./obsdiff1'))
mv.plot(area_view, diff, diff_symb)
# Extract geopoints that are hotter by 1 deg or more
#hotter = mv.filter(diff, diff >= 1)
hotter = diff.filter(diff >= 1)
# Extract geopoints that are colder by 1 deg or more
#colder = mv.filter(diff, diff <= -1)
colder = diff.filter(diff <= -1)
# Get geopoints that are within +/-1
#exact = mv.filter(diff, (diff > -1) * (diff < 1))
exact = diff.filter((diff > -1) * (diff < 1))
# Symbol visdefs for each classification
# (C) Copyright 2021 ECMWF.
#
# This software is licensed under the terms of the Apache Licence Version 2.0
# which can be obtained at http://www.apache.org/licenses/LICENSE-2.0.
# In applying this licence, ECMWF does not waive the privileges and immunities
# granted to it by virtue of its status as an intergovernmental organisation
# nor does it submit to any jurisdiction.
#
import metview as mv
from climetlab.core.ipython import ipython_active
if ipython_active:
try:
import ipywidgets # noqa
mv.setoutput("jupyter", plot_widget=True)
except ImportError:
mv.setoutput("jupyter", plot_widget=False)
def mv_read(*args, **kwargs):
return mv.read(*args, **kwargs)
def mv_read_table(*args, **kwargs):
return mv.read_table(*args, **kwargs)
land_shade = mv.mcoast(map_coastline_land_shade=True,
map_coastline_land_shade_colour="RGB(0.98,0.95,0.82)",
map_coastline_sea_shade=False,
map_coastline_sea_shade_colour="RGB(0.85,0.93,1)")
area_view = mv.geoview(map_area_definition='corners',
area=[45.83, -13.87, 62.03, 8.92],
coastlines=land_shade)
# Simplest plot:
# NOTE that when plotting a 'raw' BUFR file, Magics will plot synop symbols as shown in
# https://software.ecmwf.int/wiki/display/METV/Data+Part+1 "Plotting BUFR Data"
obs = mv.read('../tests/obs_3day.bufr')
mv.setoutput(mv.png_output(output_width=1200, output_name='./obsplot1'))
mv.plot(area_view, obs)
# ALTERNATIVELY, add an Observations Plotting visual definition
obs_plotting = mv.mobs(obs_temperature=False,
obs_cloud=False,
obs_low_cloud=False,
obs_dewpoint_colour='purple')
mv.setoutput(mv.png_output(output_width=1200, output_name='./obsplot2'))
mv.plot(area_view, obs, obs_plotting)
# ALTERNATIVELY, if we don't want to plot the whole observations, but instead want to
# extract a specific parameter from the BUFR messages, then we use the Observation Filter
# as shown here: