def num2str(n, tz=False):
"""
converts from matplotlib date format to a string
With tz=True also shows the timezone
"""
# could also use .isoformat
if tz:
return pylab.num2date(n).strftime('%Y-%m-%d %H:%M:%S.%f %Z')
else:
return pylab.num2date(n).strftime('%Y-%m-%d %H:%M:%S.%f')
def num2str(n, tz=False):
"""
converts from matplotlib date format to a string
With tz=True also shows the timezone
"""
# could also use .isoformat
if tz:
return pylab.num2date(n).strftime('%Y-%m-%d %H:%M:%S.%f %Z')
else:
return pylab.num2date(n).strftime('%Y-%m-%d %H:%M:%S.%f')
def on_key(event):
self.controller.send_message('key: {}'.format(event.key))
self.controller.event = event
if event.key == 'z':
dt = pd.to_datetime(
plt.num2date(event.xdata).strftime('%Y-%m-%d %H:%M:%S'))
self.controller.view.controlls.accordeon_start.value = dt.__str__(
)
elif event.key == 'x':
dt = pd.to_datetime(
plt.num2date(event.xdata).strftime('%Y-%m-%d %H:%M:%S'))
self.controller.view.controlls.accordeon_end.value = dt.__str__(
)
elif event.key == 'a':
self.controller.view.controlls.accordeon_alt.value = str(
event.ydata)
self._tmp_alt_x = event.xdata
def generate_xticks(self, all_days, monthsamp=1):
"""
generate xticks
all_days: array of times (days since 01-0-0)
"""
dd = pl.num2date(all_days)
xticks = []
xticklabels = []
last_month = dd[0].month
last_year = dd[0].year
xticks.append(all_days[0])
strmon = get_month_string(dd[0].month)
if self.yearonly:
xticklabels.append(str(dd[0].year))
else:
xticklabels.append(strmon + '\n' + str(dd[0].year))
for d in dd:
if (d.month != last_month) | (d.year != last_year):
xticks.append(pl.date2num(d)) # save tick location
mstr = get_month_string(d.month)
if self.yearonly:
xticklabels.append(str(d.year))
else:
if (d.year != last_year):
xticklabels.append(mstr + '\n' + str(d.year))
else:
xticklabels.append(mstr + '\n' + '')
last_month = d.month
last_year = d.year
if self.transpose:
scal = self.rescaley
else:
scal = self.rescalex
xticks = np.asarray(xticks)
# /// remap to image coordinates
ny, nx = np.shape(self.hov)
w = (xticks - pl.date2num(self.t_min)) / (pl.date2num(self.t_max)
- pl.date2num(self.t_min)) # weights for positions on x-axis
xticks = w * nx * scal
xticks = list(xticks)
# here we have monthly ticks which we can now subsample
xticks = xticks[::monthsamp]
xticklabels = xticklabels[::monthsamp]
self.x_major_locator = pl.FixedLocator(xticks)
self.x_major_formatter = pl.FixedFormatter(xticklabels)
def generate_monthly_timeseries(t, sday='01'):
"""
generate a vector monthly timeseries
t: time (numeric)
"""
tn = []
d = pl.num2date(t)
for i in range(len(t)):
y = str(d[i].year).zfill(4)
m = str(d[i].month).zfill(2)
s = y + '-' + m + '-' + sday
tn.append(pl.datestr2num(s))
return tn
def __init__(self, time, value, var_unc=None, rescalex=1,
rescaley=1, lat=None, lon=None, transpose=False,
use_cdo=True):
"""
Hovmoeller class
This class implements the functionality to generate hovmoeller
plots. It allows to calculate all relevant data by itself or
using the CDO's for calculation of e.g. zonal means
time : datetime
vector with time information
value : ndarray
data to be analyzed. if the argument lat is provided it is
assumed that lat/lon are 2D matrices
In this case the value is expected to be a 3D variables as
value(time,ny,nx)
var_unc : ndarray
additional array with variance information (can be used to
show uncertainties) - not really validated so far
rescalex : int
rescaling factor for x-variable (used to blow up the plot)
rescaley : int
rescaling factor for y-variable (used to blow up the plot)
lat : ndarray
latitude coordinates
lon : ndarray
longitude coordinates
transpose : bool
transpose the data matrix
EXAMPLES
========
#a minimum example for a time-latitude plot
#get data from e.g. file; here random data
t1=datestr2num('2011-05-01'); t2=datestr2num('2011-08-31')
d = num2date(linspace(t1,t2,1000.))
lats=linspace(-90.,90.,181) + 0.5; lats = lats[:-1]
lons = linspace(-180.,180.,361) + 0.5; lons = lons[:-1]
LONS,LATS = meshgrid(lons,lats)
ny,nx = LATS.shape; nt = len(d)
dat = rand(nt,ny,nx)
#create a hovmoeller object
myhov = hovmoeller(d,dat,lat=LATS,rescaley=20,rescalex=10)
myhov.time_to_lat(dlat=1.,yticksampling=1)
myhov.plot(title='Test Hovmoeller plot',ylabel='lat',xlabel='days',origin='lower',xtickrotation=30,climits=[-1.,1.])
show()
Another example, that does NOT calculate the hovmoeller plot by itself, but uses functionality from pyCMBS Data object
file='wachmos_sm_1978-2010_int_monthly_t63.nc'
D = Data(file,'var100',read=True)
myhov = hovmoeller(num2date(D.time),None,rescaley=20,rescalex=20)
myhov.plot(climits=[0.,0.5],input=D,xtickrotation=90,cmap='jet_r')
show()
"""
#/// check consistency ///
if value is not None:
if len(time) == len(value):
pass
else:
sys.exit('Inconsistent sizes of time and value (hovmoeller)')
if lat is not None and shape(lat) != shape(value[0, :, :]):
print shape(lat), shape(value[0, :, :])
sys.exit('Inconsistent latitudes and data (hovmoeller)')
#/// set values of class ///
self.time = time
self.transpose = transpose
ntim = len(self.time)
t = pl.date2num(time)
self.t_min = pl.num2date(np.floor(t.min()))
self.t_max = pl.num2date(np.ceil(t.max()))
if value is not None:
self.value = value.copy()
self.value = ma.array(self.value, mask=isnan(self.value))
# now reshape everything to [time,ngridcells]
self.value.shape = (ntim, -1)
if var_unc is None:
self.var_unc = None
else:
self.var_unc = var_unc.copy()
self.var_unc.shape = (ntim, -1)
self.hov_var = None
self.rescalex = rescalex
self.rescaley = rescaley
if lat is not None:
self.lat = lat.copy()
self.lat.shape = (-1)
else:
self.lat = None
if lon is not None:
self.lon = lon.copy()
self.lon.shape = (-1)
else:
self.lon = None
self.hov = None
def dateplt1(t,y,add=None,yl=5,fxlim=None,fylim=None,fighold=False,nomth=False,**kwargs):
from matplotlib.dates import YearLocator,DateFormatter,MonthLocator
if add is None:
ymin = np.nanmin(y) - 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
ymax = np.nanmax(y) + 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
if isinstance(y,np.ma.MaskedArray):
data_extent = np.where(y.mask == False)
if len(data_extent) == 0:
data_start = 0
data_end = y.size - 1
else:
data_start = data_extent[0][0]
data_end = data_extent[0][-1]
else:
data_start = 0
data_end = y.size - 1
if (isinstance(t[0],np.float64) is True) or (isinstance(t[0],np.float32) is True):
datemin = num2date(t[data_start])
datemax = num2date(t[data_end])
elif isinstance(t[0],dt):
datemin = t[data_start]
datemax = t[data_end]
else:
return None
datediff = datemax - datemin
datediff = datediff / 50
datemin = datemin - datediff
datemax = datemax + datediff
if fighold is True:
plt.hold(False)
fig = plt.gcf()
plt.clf()
else:
fig = plt.figure()
if (fylim is not None):
if type(fylim) is not tuple:
print "option <fylim> must be a tuple"
return None
(ymin,ymax) = fylim
ax = fig.add_subplot(111,autoscale_on=False,ylim=(ymin,ymax))
if (np.sign(ymax) != np.sign(ymin)):
ax.axhline(0.0,c='black',lw=1.2,alpha=0.7)
line1 = ax.plot(t,y,'+-',**kwargs)
years = YearLocator(yl)
yearsFmt = DateFormatter('%Y')
if nomth == False:
months = MonthLocator(1)
ax.set_xlim(datemin,datemax)
#fig.autofmt_xdate()
ax.xaxis.set_major_locator(years)
ax.xaxis.set_major_formatter(yearsFmt)
if nomth == False:
ax.xaxis.set_minor_locator(months)
ax.grid(True)
my_fmt_xdate(ax)
##plt.show()
##ylabel('temperature anomaly')
plt.figure(fig.number)
return line1
else:
xchange = 0
ychange = 0
fig = plt.gcf()
ax = plt.gca()
years = YearLocator(yl)
yearsFmt = DateFormatter('%Y')
months = MonthLocator(1)
line1 = ax.plot(t,y,'+-',**kwargs)
if isinstance(y,np.ma.MaskedArray):
data_extent = np.where(y.mask == False)
if len(data_extent) == 0:
data_start = 0
data_end = y.size - 1
else:
data_start = data_extent[0][0]
data_end = data_extent[0][-1]
else:
data_start = 0
data_end = y.size - 1
if (isinstance(t[0],np.float64) is True) or (isinstance(t[0],np.float32) is True):
datemin = num2date(t[data_start])
datemax = num2date(t[data_end])
elif isinstance(t[0],dt):
datemin = t[data_start]
datemax = t[data_end]
else:
return None
datediff = datemax - datemin
datediff = datediff / 50
datemin = datemin - datediff
datemax = datemax + datediff
ymin = np.nanmin(y) - 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
ymax = np.nanmax(y) + 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
(oldymin,oldymax) = ax.get_ylim()
(oldxmin,oldxmax) = ax.get_xlim()
if (fylim is not None):
if type(fylim) is not tuple:
print "option <fylim> must be a tuple"
return None
(ymin,ymax) = fylim
if datemin.toordinal() < oldxmin:
xchange = 1
oldxmin = datemin.toordinal()
if datemax.toordinal() > oldxmax:
xchange = 1
oldxmax = datemax.toordinal()
if ymin < oldymin:
ychange = 1
oldymin = ymin
if ymax > oldymax:
ychange = 1
oldymax = ymax
if xchange:
plt.setp(ax,xlim=ax.set_xlim(oldxmin,oldxmax))
if ychange:
plt.setp(ax,ylim=ax.set_ylim(oldymin,oldymax))
ax.xaxis.set_major_locator(years)
ax.grid(True)
my_fmt_xdate(ax)
plt.figure(fig.number)
return line1
def dateplt_m2(t,y,yl=5,fxlim=None,fylim=None,ticksize=10,nomth=False,fmt='+-',**kwargs):
"""Different from dateplt_m in that it's just a 2-digit year for x tick marks"""
from matplotlib.dates import YearLocator,DateFormatter,MonthLocator
xchange = 0
ychange = 0
fig = plt.gcf()
ax = plt.gca()
years = YearLocator(yl)
yearsFmt = DateFormatter("'%y")
if nomth == False:
months = MonthLocator(1)
if isinstance(y,np.ma.MaskedArray):
data_extent = np.where(y.mask == False)
if len(data_extent) == 0:
data_start = 0
data_end = y.size - 1
else:
data_start = data_extent[0][0]
data_end = data_extent[0][-1]
else:
data_start = 0
data_end = y.size - 1
if (isinstance(t[0],np.float64) is True) or (isinstance(t[0],np.float32) is True):
datemin = num2date(t[data_start])
datemax = num2date(t[data_end])
elif isinstance(t[0],dt):
datemin = t[data_start]
datemax = t[data_end]
else:
return None
datediff = datemax - datemin
datediff = datediff / 50
datemin = datemin - datediff
datemax = datemax + datediff
ymin = np.nanmin(y) - 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
ymax = np.nanmax(y) + 0.02 * (abs(np.nanmax(y) - np.nanmin(y)))
(oldymin,oldymax) = ax.get_ylim()
(oldxmin,oldxmax) = ax.get_xlim()
if (np.sign(ymax) != np.sign(ymin)):
ax.axhline(0.0,c='black',lw=1.2,alpha=0.7)
line1 = ax.plot(t,y,fmt,**kwargs)
# ax.set_xlim(datemin,datemax)
ax.grid(True)
if (datemin.toordinal() < oldxmin) | (oldxmin == 0.):
xchange = 1
oldxmin = datemin.toordinal()
if datemax.toordinal() > oldxmax:
xchange = 1
oldxmax = datemax.toordinal()
if ymin < oldymin:
ychange = 1
oldymin = ymin
if ymax > oldymax:
ychange = 1
oldymax = ymax
if (fylim is not None):
if type(fylim) is not tuple:
print "option <fylim> must be a tuple"
return None
(ymin,ymax) = fylim
plt.setp(ax,ylim=ax.set_ylim(ymin,ymax))
else:
if xchange:
plt.setp(ax,xlim=ax.set_xlim(oldxmin,oldxmax))
if ychange:
plt.setp(ax,ylim=ax.set_ylim(oldymin,oldymax))
ax.xaxis.set_major_locator(years)
ax.xaxis.set_major_formatter(yearsFmt)
if nomth == False:
ax.xaxis.set_minor_locator(months)
xtl = ax.get_xticklabels()
ytl = ax.get_yticklabels()
plt.setp(xtl,'size',ticksize)
plt.setp(ytl,'size',ticksize)
##plt.show()
##my_fmt_xdate(ax,rot=45,hal='right')
plt.figure(fig.number)
return line1
wind_gust1_f2 = buoy.variables['wind_gust1_f2'][:] pressure = buoy.variables['pressure'][:] wave_h_max = buoy.variables['wave_h_max'][:] wind_dir1_f2 = buoy.variables['wind_dir1_f2'][:] dew_point = buoy.variables['dew_point'][:] avg_wind_int1_f2 = buoy.variables['avg_wind_int1_f2'][:] cm_dir3 = buoy.variables['cm_dir3'][:] cm_dir2 = buoy.variables['cm_dir2'][:] longitude = buoy.variables['longitude'][:] avg_wind_int1 = buoy.variables['avg_wind_int1'][:] avg_wind_int2 = buoy.variables['avg_wind_int2'][:] wind_gust2 = buoy.variables['wind_gust2'][:] avg_dir2 = buoy.variables['avg_dir2'][:] wind_gust1 = buoy.variables['wind_gust1'][:] #converte data em juliano (time) para datetime datat = pl.num2date(time) #figuras pl.figure() pl.subplot(311) pl.plot(datat,wave_hs,'.') pl.ylim(0,7) pl.subplot(312) pl.plot(datat,wave_period,'.') pl.ylim(0,25) pl.subplot(313) pl.plot(datat,wave_dir,'.') pl.ylim(0,360) pl.show()
def get_timed_data_sheet(wb, fkeys, pkeys, date_nums, ras, decs, azs, elevs):
"""
See if the timed data sheet exist and create it if not
A timed data sheet is basically the power VtoF data as a function
of time (rows) and frequency/polarization (columns). This initiates
the sheet but does not fill it. fill_map_sheet() does that.
@param wb : the spreadsheet
@type wb : openpyxl.Wotkbook() instance
@param fkeys : frequency-based keys into the data
@type fkeys : list of float
@param pkeys : polarization-based keys
@type pkeys : list of str
@param date_nums : datetime for each sample
@type date_nums : matplotlib date numbers
@param ras : right ascensions
@type ras : list of floats (hr)
@param decs : declinations
@type decs : list of floats (deg)
@param azs : azimuths
@type azs : list of floats
@param elevs : elevations
@type elevs : list of float
@return: datasheet
"""
timedata_sheet = wb.get_sheet_by_name("TimeData")
if timedata_sheet == None:
# Doesn't exist. Make it.
if diag:
print("Trying to create a new data sheet")
try:
timedata_sheet = wb.create_sheet(1)
except Exception as details:
print("Could not create sheet")
print(details)
return None
timedata_sheet.title = "TimeData"
if diag:
print(timedata_sheet,"has been given the name",\
timedata_sheet.title)
# make column titles
column_names = ["Date/Time", "R.A.", "Decl.", "Az", "Elev"]
for freq in fkeys:
for pol in pkeys:
column_names.append(("%5.2f-%3s" % (freq, pol)).strip())
for index in range(len(column_names)):
timedata_sheet.cell(row=0,
column=index).value = column_names[index]
# Fill the columns
for count in range(len(date_nums)):
timedata_sheet.cell(row = count+1,
column = get_column_id(timedata_sheet,"Date/Time")).value = \
num2date(date_nums[count])
timedata_sheet.cell(row = count+1,
column = get_column_id(timedata_sheet,"R.A.")).value = \
ras[count]
timedata_sheet.cell(row = count+1,
column = get_column_id(timedata_sheet,"Decl.")).value = \
decs[count]
timedata_sheet.cell(row = count+1,
column = get_column_id(timedata_sheet,"Elev")).value = \
elevs[count]
timedata_sheet.cell(row = count+1,
column = get_column_id(timedata_sheet,"Az")).value = \
azs[count]
if diag:
print(timedata_sheet.title, "sheet exists")
return timedata_sheet
