""" canam4_prep.py:

9/9/2014: Goal is to prep the data for 1. retrieval/calcs and 2. plotting

To be used in conjunction with simulation_funcs.py

Splitting canam4sims_analens.py

This is particular to canam4, but the functions being called

should not be (in the end...) so can have prep scripts for CanESM2, etc

"""

import constants as con

import simulation_funcs as sfnc

import statshelper as sh

sh=reload(sh)

sfnc = reload(sfnc)

con = reload(con)

#ccm = reload(ccm)

plt.close("all")

plt.ion()

# Will use this script to set up all the stuff.

# for example, plot_seasonal_maps() requires:

# simpair keys

# fielddict with 'field','ncfield','fieldstr'

# pparams dict with cmin, cmax, cmap, latlim, type, suppcb ....

# if vert=True, also screen, levlim, addcontlines

# coords = {'lat': con.get_t63lat(), 'lon': con.get_t63lon()}

printtofile=False

field = 'st'

smclim=True

level=50000 # for threed

addcont=False # overlay map with contours

sigoff=True # if True, don't add significance

effdof=False # use effective deg of freedom or no. Set to False.

#field2='sicn'

field2='gz'

#field2='pmsl'

level2=50000

# seasonalmap, seasonalvert, plotzonmean, plotseacyc, pattcorrwithtime, plotregmean,calcregmeanwithtime, calcregunccascade,calccomposites, timetosig, timetosigsuper, plotscatter

plottype='seasonalmap'

projtype='nh' #'eastere' # 'nh','sh','sq','eastere','nastere','eabksstere','ealamb','eabkslamb'

# None, nh, polcap60, polcap65, polcap70, eurasia, eurasiamori, eurasiasth,eurasiathin,eurasiathinw,eurasiathine,ntham, nthatl, bks, bksmori, soo

region='polcap60' #'eurasiamori' #'polcap60' #'eurasia' #'eurasiamori'

screen=True # just for seasonalvert

seacyclatlim=60

withlat=False

pattcorryr=False # need anymore?

latlim = 45 # None #45 # lat limit for NH plots. Set to None otherwise. use 45 for BC-type maps

round=False # if latlim is not None, this specifies whether the polar stereographic figure should be square or round

levlim= 100 # level limit for vertical ZM plots (in hPa). ignored if screen=True

fallwin=False # just SON and DJF

bimon=False # do bi-montly seasons instead

figtrans=False # for maps/vert: make seasons the cols and sims the rows if True. auto True for simsforpaper

# Choose how to handle the data ==============

normbystd=False

pct = False # if True, do calculation as a percent (@@not fully implemented)

halftime=False # get only the first 60yrs. make sure to set the other flag the opp

halftime2=False # get only the last 60yrs. make sure to set the other flag the opp

# Choose what simulations to add =============

# default is R1-5, ENS

canens=False # just the CAN ensemble (E1-E5) plus mean, plus mean of R ensemble. option to addobs only.

allens=False # this is ONLY the ensemble means, plus superensemble

sensruns=False # sensruns only: addr4ct=1,addsens=1. others=0 no meanBC, r mean, or obs

ivar=False # this will show ENS (TOT) and ENSE (ANTH) and their difference = internal var

simsforpaper=False # ANT, TOT, HAD, NSIDC only. best for maps and zonal mean figs (not line plots)

simsforpaperwace=False # Just R1,R5 (warm/cold Eurasia) or whatever it's set to below.

antcat=False # this is the concatenation of ens members within each ensemble (really only useful for ANT)

bothcat=False # can do concatenation of both ensembles if want to. These are useful for timetosig

onlyens=False # just do ensemble means ANT and TOT

addobs=True # add mean of kemhad* & kemnsidc* runs to line plots, seasonal maps.

addr4ct=False # add kem1pert2r4ct (constant thickness version of ens4)

addsens=False # add sensitivity runs (kem1pert1b, kem1pert3)

addrcp=False # add kem1rcp85a simulation (and others if we do more)

addcanens=False # add "initial condition" ensemble of kemctl1/kem1pert2

addsuper=False # add superensemble mean

# not sure these flags are in use?

sigtype = 'cont' # significance: 'cont' or 'hatch' which is default

siglevel= 0.10 # 0.10, 0.05

print 'SIGLEVEL = ' + str(siglevel) # @@@

ptparams={}

ptparams['smclim'] = smclim # for maps typically

ptparams['latlim'] = latlim

ptparams['round'] = round

ptparams['levlim'] = levlim

ptparams['region'] = region # plotregmean

ptparams['screen'] = screen # seasonalvert

ptparams['seacyclatlim'] = seacyclatlim # plotseacyc

ptparams['withlat'] = withlat # plotseacyc

ptparams['pattcorryr'] = pattcorryr # pattcorrwithtime

if plottype=='seasonalvert':

seasonalvert=True

else:

seasonalvert=False

## # Choose type of plot =========================

## seasonalmap=False # seasonal maps (SON, DJF, MAM, JJA)

## seasonalvert=False # seasonal vertical zonal means instead of maps

## screen=False # whether to have screen-style vertical zonal means

## plotzonmean=False # plotzonmean,plotseacyc,pattcorrwithtime are mutually exclusive

## plotseacyc=False # plotzonmean,plotseacyc,pattcorrwithtime are mutually exclusive

## seacyclatlim=60 # southern limit for plotting polar mean seasonal cycles (line plot)

## withlat=False # plot the seasonal cycle with latitude dimension too (only for plotseacyc=1)@@for now just std over ens

## #squatseacyc=False # plot seacycle figs as shorter than wide

## #squatterseacyc=True # even shorter, for paper

## pattcorrwithtime=False # plot pattern correlation with time for each ens member

## pattcorryr=False # if True, do a yearly anomaly pattern rather than time-integrated

## plotregmean=True

## region = 'ntham' # None, polcap60, polcap65, polcap70, eurasia, eurasiamori, ntham, nthatl, bks, bksmori, soo

testhadisst=0 # check which ens member most similar to hadisst

model='CanAM4'

# #################################################################

# Probably don't need to modify below if everything goes well.

# #################################################################

# initialize things that don't get set otherwise

savestr='' # string for plot filenames

## threed=False # is the field three dimensional

## sia=False # is the requested field sea ice area

## conv=1

timesel=None

## isflux=False

shadeens=('histBC',)

corrlim=45 # southern lat limit for pattern correlation with time

# set up simulations and figure filename strings

sims = 'R1','R4','R3','R5','R2','ENS'#,'ENSE'#,'CAN' # R's in order of sea ice loss

defaultsims=sims

seasons = ('SON','DJF','MAM','JJA')

biseas = ('SO','ND','JF') # @@@ so far only these implemented. expecting to add all 11/25/14

#biseas = ('ND',) # @@@@

if simsforpaper: # best for maps only

sims = ('HAD','NSIDC','ENSE','ENS')

savestr = '_forpap4' # add ENS. # 4 means fig is transposed

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

else:

seasons=('SON','DJF')

figtrans=True

elif simsforpaperwace:

#print '@@@ simsforpaperwace is WACE paper -- ND cold and warm extremes, R1, R5 only'

#savestr = '_forpapwace'; sims = ('R1','R5'); seasons=('DJF',); figtrans=True # 95% stat sig

#print '@@@ simsforpaperwace (version d) is WACE paper -- DJF cold and warm extremes, E4,R4 only'

#savestr = '_forpapwaced'; sims = ('E4','R4'); seasons=('DJF',); figtrans=True

#print '@@@ simsforpaperwace (version e) is WACE paper -- DJF cold and warm extremes, E4,E1 only'

#savestr = '_forpapwaceE'; sims = ('E4','E1'); seasons=('DJF',); #figtrans=True

#print '@@@ simsforpaperwace (version c) is WACE paper -- DJF cold and cold extremes, R2, E4 only'

#savestr = '_forpapwacec'; sims = ('R2','E4'); seasons=('DJF',); figtrans=True

#print '@@@ simsforpaperwace is WACE paper -- ND ENSEMBLE MEANS ONLY'

#savestr = '_forpapwace_ensmean'; sims = ('ENS','ENSE'); seasons=('ND',); figtrans=True # 95% stat sig

#print '@@@ simsforpaper is WACE B paper right now -- ND cold and warm extremes, E4, E1 only'

#savestr = '_forpapwaceb'; sims = ('E4','E1'); seasons=('ND',); figtrans=False # 90% stat sig

print '@@@ simsforpaperwace (version F) is WACE paper -- DJF sicn for NSIDC and ANT'

savestr = '_forpapwaceF'; sims = ('NSIDC','ENSE'); seasons=('DJF',); #figtrans=True

elif sensruns: # add sensitivity runs. with Shaded ENS. don't plot meanBC, mean of ens

sims = sims[0:5] + ('R4ct','CANnosst','CANnothk') # @@ change to E1nosst, etc?

savestr = '_sensruns'

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

elif canens: # do canens instead of r ens. Useful for maps.

sims = ('E1','E2','E3','E4','E5','ENSE','ENS')

savestr = '_canensonly'

shadeens=('histIC',)

if addobs:

sims = sims + ('HAD','NSIDC')

savestr = savestr + 'obs'

if addsuper:

sims = sims + ('ESPR',)

savestr = savestr + 'spr'

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

elif allens: # just do the ens means and superensemble mean

sims = ('ENS','ENSE','ESPR')

savestr = '_allens'

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

# addobs? @@@

elif ivar:

sims = ('ENS','ENSE','IVAR')

savestr = '_ivar'

smclim=True # @@

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

elif antcat: # eventually add totcat @@@

sims = ('ENSECAT',)

savestr = '_antcat'

figtrans=True

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

elif bothcat: # both antcat and totcat

sims = ('ENSECAT','ENSCAT')

savestr = '_anttotcat'

figtrans=True

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

elif onlyens:

sims = ('ENSE','ENS')

savestr = '_onlyens'

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

else:

savestr=savestr+'_ens' # default sims

if addcanens:

sims = sims + ('E1','E2','E3','E4','E5','ENSE') # E1=CAN

savestr = savestr + 'canens'

shadeens=shadeens+('histIC',)

else:

sims = sims + ('ENSE',)

if addsuper:

sims = sims + ('ESPR',)

savestr = savestr + 'spr'

if addobs:

sims = sims + ('HAD','NSIDC')

savestr = savestr + 'obs'

if addr4ct:

sims = sims + ('R4ct',)

savestr = 'r4ct' # for figure filenames

if addsens:

sims = sims + ('CANnosst','CANnothk') # control is kemctl1 (or '' key)

savestr = savestr + 'sens'

if addrcp:

sims = sims + ('RCPa',) # control is kemctl1

savestr = savestr + 'rcpa'

if bimon:

seasons=biseas

savestr = savestr + 'bimon'

if fallwin:

seasons=('SON','DJF')

savestr = savestr + '_SONDJF'

import load_fldmeta as ld

ld=reload(ld)

## ############################ start copy to load_fldmeta.py

## # # # ######## set Field info ###################

## # gz, t, u, v, q (3D !)

## # st, sic, sicn (sia), gt, pmsl, pcp, hfl, hfs, turb, net, flg, fsg, fn, pcpn, zn, su, sv (@@later ufs,vfs)

## print field

if halftime:

timesel= '0002-01-01,0061-12-31'

elif halftime2:

timesel='0062-01-01,0121-12-31'

## # # # ###########################################

## # Shouldn't have to mod below....

## """ plev = 100, 200, 300, 500, 700, 1000, 2000, 3000, 5000, 7000, 10000, 12500,

## 15000, 17500, 20000, 22500, 25000, 30000, 35000, 40000, 45000, 50000,

## 55000, 60000, 65000, 70000, 75000, 77500, 80000, 82500, 85000, 87500,

## 90000, 92500, 95000, 97500, 100000 ;

## """

## fdict = {'field': field, 'ncfield': None, 'fieldstr': None,

## 'units': None, 'conv': conv,

## 'nonstandardlev': nonstandardlev,

## 'threed': threed} # fielddict

## # reserved for expansion into the plotfunction call

## pparams = {'cmin': None, 'cmax': None, 'cmap': 'blue2red_20',

## 'type':'nh', 'latlim': latlim} # plotparams

seacycylim=None

infodict ={'cmapclimo': 'Spectral_r','leglocs': None,

'ptype': projtype,'contclr': '0.3', 'contthk': 1, 'contstl': None} # random other info. projtype for maps only

fdict,pparams=ld.loadfldmeta(field,infodict,plottype,ptparams,level=level)

if addcont: # overlay map with another field in contours

# start with just anomaly contours. @@later add option for climo

fdict2,pparams2=ld.loadfldmeta(field2,infodict,plottype,ptparams,level=level2)

coords = {'lev': con.get_t63lev(), 'lat': con.get_t63lat(), 'lon': con.get_t63lon()}

## infodict['savestr'] = savestr

## infodict['leglocs'] = leglocs

## infodict['seacycylim'] = seacycylim

## fdict['isflux'] = isflux

## fdict['threed'] = threed

## ###################### end copy to load_fldmeta.py

# do an if elif elif ....

if plottype=='seasonalvert':

infodict['screen']=screen

else:

infodict['screen'] = None

if plottype in ('seasonalmap','seasonalvert'):

print fdict

print sims

print pparams

#pparams2['cmin']=-.6; pparams2['cmax']=.6 #@@@

# this one does data processing and plotting together

# some stuff in the function need to be removed or set differently.@@

# marked in the function @@

if addcont:

addflds=(fdict2,)

if simsforpaperwace and field2=='sicn':

pparams2['cmin'] = -.50; pparams2['cmax'] = 0 # only plot negative contours??

infodict['contclr'] = 'w'

infodict['contthk'] = 2 # thickness of contour lines

infodict['contstl'] = '-' # solid linestyles

addpparams=(pparams2,)

else:

addflds=None

addpparams=None

thefig = sfnc.calc_plot_seasonal_maps(fdict,coords,sims,pparams,vert=seasonalvert,

loctimesel=timesel,info=infodict,seas=seasons,

printtofile=printtofile,addflds=addflds,addpparams=addpparams)

if simsforpaperwace:

# here I will adjust the figure for the WACE paper

if figtrans:

if projtype in ('eastere','ealamb'):

thefig.set_size_inches((5,8))

thefig.subplots_adjust(hspace=.02,wspace=.02)

elif projtype in ('eabksstere','eabkslamb'):

thefig.set_size_inches((4,8.5))

thefig.subplots_adjust(hspace=.02,wspace=.02)

else:

thefig.set_size_inches((4,10))

else:

thefig.set_size_inches((10,5))

theaxs = thefig.get_axes()

ax1=theaxs[0]

"""if field2=='sicn':

fdict2,pparams2=ld.loadfldmeta(field2,infodict,plottype,ptparams,level=level2)

dblob2 = sfnc.calc_seasons(fdict2,coords,sims,seas=seasons,loctimesel=timesel,info=infodict,effdof=effdof)

fldctl=dblob2['ctl'][sims[0]]['DJF']

fldpt=dblob2['pert'][sims[0]]['DJF']

# @@@@@ ADD SICN 15% (or all?) contours here (make them nice:

# thick and light colored over the dark red)

lons,lats = np.meshgrid(coords['lon'],coords['lat'])

#figtmp,axtmp=plt.subplots(1,1)

# this needs to be bm.contour() @@

ax1.contour(lons,lats,fldctl,colors='0.6',linewidths=1)#,levels=[0,0.15,0.15]"""

if 'ENS' in sims: # assume it is just the two ensemble means

prstr='ANTTOT'

ax1.set_title('a. Individual SIC forcings')

ax1.set_ylabel('')

elif ('R2' in sims) and ('DJF' in seasons): # assume it's the two cooling cases for DJF

prstr='c'

ax1.set_title('a. Individual SIC forcings')

ax1.set_ylabel('Latitude',fontsize=12)

if not figtrans:

ax1.set_xlabel('Longitude',fontsize=12)

elif (field=='sicn') and ('NSIDC' in sims):

prstr=''

ax1.set_title('a. NSIDC SIC forcing')

ax1.set_ylabel('',fontsize=12)

if not figtrans:

ax1.set_xlabel('',fontsize=12)

else: # else assume it's the cooling and warming cases for paper

prstr=''

#ax1.set_title('a. Cooling case')

ax1.set_title('c. Cooling case')

ax1.set_ylabel('Latitude',fontsize=12)

if not figtrans:

ax1.set_xlabel('Longitude',fontsize=12)

ax2=theaxs[1]

if 'ENS' in sims: # assume it is just the two ensemble means

ax2.set_title('b. Average SIC forcing')

ax2.set_ylabel('')

elif ('R2' in sims) and ('DJF' in seasons): # assume it's the two cooling cases for DJF

prstr='c'

ax2.set_title('b. Average SIC forcing')

ax2.set_xlabel('Longitude',fontsize=12)

if figtrans:

ax2.set_ylabel('Latitude',fontsize=12)

else:

ax2.set_ylabel('',fontsize=12)

elif (field=='sicn') and ('NSIDC' in sims):

ax2.set_title('b. Average SIC forcing')

ax2.set_xlabel('',fontsize=12)

if figtrans:

ax2.set_ylabel('',fontsize=12)

else:

ax2.set_ylabel('',fontsize=12)

else: # else assume it's the cooling and warming cases for paper

if 'DJF' in seasons:

prstr='d' # version d

#ax2.set_title('b. Warming case')

ax2.set_title('d. Warming case')

ax2.set_xlabel('Longitude',fontsize=12)

if figtrans:

ax2.set_ylabel('Latitude',fontsize=12)

else:

ax2.set_ylabel('',fontsize=12)

if not figtrans: # one row

axcb=theaxs[2] # colorbar

axcb.set_position([.91,.25, .02,.5])

else: # one col

axcb=theaxs[2] # colorbar

axcb.set_position([.15,0.07, 0.75, .02])

if 'ENS' not in sims:

axcb.set_xticks(np.arange(-1,1.2,0.2))

axcb.set_xticklabels(('-1.0','','','','','0','','','','','1.0'))

thefig.suptitle('')

if figtrans:

trans='trans_'

else:

trans=''

if addcont:

thefig.savefig('wacefigure4'+savestr+'_' + field + '_' + field2 + 'cont_' + trans + prstr + projtype + '.pdf')

thefig.savefig('wacefigure4'+savestr+'_' + field + '_' + field2 + 'cont_' + trans + prstr + projtype + '.eps')

else:

thefig.savefig('wacefigure4'+savestr+'_' + field + '_' + trans + prstr + projtype + '.pdf')

thefig.savefig('wacefigure4'+savestr+'_' + field + '_' + trans + prstr + projtype + '.eps')

if plottype=='plotseacyc':

dblob = sfnc.calc_seasonal_cycle(fdict,coords,sims,withlat=withlat,loctimesel=timesel,info=infodict,siglevel=siglevel)

sfnc.plot_seasonal_cycle(dblob,fdict,sims,ptypes=('anom','stddev'),info=infodict,printtofile=printtofile)

if plottype=='plotzonmean':

dblob = sfnc.calc_seasons(fdict,coords,sims,loctimesel=timesel,info=infodict,calctype='zonmean',effdof=effdof,siglevel=siglevel)

sfnc.plot_zonmean_byseas(dblob,fdict,coords,sims,ptypes=('climo','anom','stddev','stdan'),info=infodict,printtofile=printtofile)

if plottype=='pattcorrwithtime':

if pattcorryr==True:

calctype='pattcorrwithtimeyr'

else:

calctype='pattcorrwithtime'

dblob = sfnc.calc_seasons(fdict,coords,sims,loctimesel=timesel,info=infodict,calctype=calctype,effdof=effdof,siglevel=siglevel)

sfnc.plot_pattcorrwithtime_byseas(dblob,fdict,sims,info=infodict,calctype=calctype,printtofile=printtofile)

if plottype=='plotregmean':

dblob = sfnc.calc_seasons(fdict,coords,sims,seas=seasons,loctimesel=timesel,info=infodict,calctype='regmean',effdof=effdof,siglevel=siglevel)

sfnc.plot_regmean_byseas(dblob,fdict,sims,info=infodict,seas=seasons,printtofile=printtofile)

# DEBUG:

ci=dblob['ci']

pv=dblob['pval']

import pandas as pd

# calc whether ens means are sig different @@@@

if shadeens != None: # right now, will never be none. Should change that. @@@

# calc anom range vs mean anom for the given field within given ensemble

allensdt,allensmdt = con.build_ensembles(shadeens,dblob,calctype='diff')

for ens in shadeens:

ensdf=pd.DataFrame(allensdt[ens])

#ensmdf=pd.DataFrame(allensmdt[ens])

ensrng = ensdf.max(axis=1)-ensdf.min(axis=1)

ensrat = ensrng / ensdf.mean(axis=1)

print ens + ' RANGE: '

print str(ensrng)

print ens + ' MEAN: '

print str(ensdf.mean(axis=1))

print ens + ' RATIO*100: '

print str(ensrat*100)

if len(shadeens)>1:

sh.calc_ensemblestats(dblob,shadeens,seas=seasons)

if plottype=='timetosig' or plottype=='timetosigsuper':

calctype='timetosig'

pparams['cmin'] = 0; pparams['cmax'] = 120 # same clims for all runs/vars in this project

pparams['cmap'] = 'Spectral_r' #'YlGnBu'

if plottype=='timetosigsuper':

calctype='timetosigsuper'

pparams['cmax'] = 600

dblob = sfnc.calc_seasons(fdict,coords,sims,loctimesel=timesel,info=infodict,

calctype=calctype,seas=seasons,effdof=effdof,siglevel=siglevel)

sfnc.plot_seasonal_maps(dblob,fdict,coords,sims,pparams,plottype='timetosig',

vert=False,seas=seasons,info=infodict,printtofile=printtofile)

if plottype=='calcregmeanwithtime' or plottype=='calcregunccascade' or plottype=='calccomposites':

dblob = sfnc.calc_seasons(fdict,coords,sims,seas=seasons,

loctimesel=timesel,info=infodict,calctype='regmeanwithtime',

effdof=effdof,siglevel=siglevel)

infodict['region'] = 'eurasiathicke'; #'eurasiathine' #'eurasiamori'

rg2 = infodict['region']

dblob2 = sfnc.calc_seasons(fdict,coords,sims,seas=seasons,

loctimesel=timesel,info=infodict,calctype='regmeanwithtime',

effdof=effdof,siglevel=siglevel)

import pandas as pd

# dblob should have regional means *with* time dimension

# want to test sig different mean and variance b/w ctl and pert

allstats=sh.calc_runstats(dblob,sims, seas=seasons,siglevel=siglevel)

fpvaldf=pd.DataFrame(allstats['fpval']) # pval of f statistic for variance significance

fpvaldft=fpvaldf.transpose()

tpvaldf=pd.DataFrame(allstats['tpval']) # pval of t statistic for mean significance

tpvaldft=tpvaldf.transpose()

import cccmacmaps as ccm

if plottype=='calcregunccascade':

effdof=False

col=('0.3',ccm.get_linecolor('firebrick'))

if field=='st' and region in ('eurasia','eurasiamori'):

#xlab = '$\Delta$ Eurasia SAT ($^\circ$C)'

xlab = '($^\circ$C)'

elif field=='st' and region=='polcap60':

#xlab = '$\Delta$ >60$^\circ$N SAT ($^\circ$C)'

xlab = '($^\circ$C)'

else:

xlab=None

for sea in seasons:

seas=(sea,)

xlims=None

fig,axs = plt.subplots(2,1)

fig.set_size_inches(4,8)

fig.set_frameon(False)

ax=axs[0] # first polar cap SAT

if 'ND' in seas:

annlab='a. Nov-Dec Polar SAT changes'

elif 'DJF' in seas:

annlab='a. Dec-Jan-Feb Polar SAT changes'

xlims=(0.4,1.9) # for polar cap (top cascade)

else:

annlab='a. Polar SAT changes'

annloc=(0.02,0.94)

sfnc.plot_uncertainty_cascade(dblob,fdict,coords,sims,pparams,

info=infodict,seas=seas,ax=ax,xlab=xlab,annlab=annlab,annloc=annloc,

color=col,effdof=effdof,xlims=xlims,siglevel=siglevel)

fs=9

if 'ND' in seas:

ax.annotate('Ensemble mean',xy=(1.9,2.93),xycoords='data',fontsize=fs)

ax.annotate('120-yr means', xy=(1.9,1.93),xycoords='data',fontsize=fs)

ax.annotate('60-yr means', xy=(1.9,0.93),xycoords='data',fontsize=fs)

ax.annotate('Individual SIC forcings',xy=(1.31,6.25),xycoords='data',fontsize=fs)

ax.annotate('Average SIC forcing',xy=(1.31,3.25),xycoords='data',fontsize=fs)

annlab='b. Nov-Dec Eurasian SAT changes'

elif 'DJF' in seas:

xlims=(-.4,.65) # for second cascade

ax.annotate('Ensemble mean',xy=(1.4,2.93),xycoords='data',fontsize=fs)

ax.annotate('120-yr means', xy=(1.4,1.93),xycoords='data',fontsize=fs)

ax.annotate('60-yr means', xy=(1.4,0.93),xycoords='data',fontsize=fs)

ax.annotate('Individual SIC forcings',xy=(0.45,6.25),xycoords='data',fontsize=fs)

ax.annotate('Average SIC forcing',xy=(0.45,3.25),xycoords='data',fontsize=fs)

if rg2 in ('nthamsth','ntham'):

annlab = 'b. Dec-Jan-Feb N. American SAT changes'

xlims=(-.6,.5)

elif rg2 in ('eurasiathicke',):

annlab='b. Dec-Jan-Feb Eurasian SAT changes'

xlims=(-.4,.75)

else:

annlab='b. Dec-Jan-Feb Eurasian SAT changes'

else:

annlab='b. Eurasian SAT changes'

annloc=(0.02,0.94)

ax=axs[1] # second eurasian SAT

sfnc.plot_uncertainty_cascade(dblob2,fdict,coords,sims,pparams,

info=infodict,seas=seas,ax=ax,xlab=xlab,annlab=annlab,annloc=annloc,

color=col,effdof=effdof,xlims=xlims,siglevel=siglevel)

if printtofile:

if effdof:

fig.savefig(field + '_' + region + '_' + seas[0] + '_unccascade3_effdofobs.pdf')

else:

fig.savefig(field + '_polcap60' + rg2 + '_' + seas[0] + '_unccascade7ann.pdf')

#fig.savefig(field + '_polcap60' + rg2 + '_' + seas[0] + '_unccascade7ann.eps')

elif plottype=='calccomposites':

print 'here calculate high and low composites using a given threshold, say 1 sigma' # @@@@

thresh=1 # the multiplier to sigma, ie 1sigma or 2sigma etc.

# first do eurasia SAT anomalies.

# assume I have a timeseries of anomalies, with a sigma associated with it.

# select all times with val < sigma and with val > sigma

#diffdf = pd.DataFrame(dblob['diff'])

#std= diffdf.std()

diffdt=dblob['diff']

for sea in seasons:

for skey in sims:

diff = diffdt[skey][sea]

std = diff.std()

maxthresh = diff.mean()+thresh*std

minthresh = diff.mean()-thresh*std

high = ma.masked_where(diff<maxthresh,diff)

low = ma.masked_where(diff>minthresh,diff)

xx=np.arange(0,len(diff))

fig,axs=plt.subplots(1,2)

ax=axs[0] # timeseries

ax.plot(xx,diff,color='k')

ax.axhline(y=0,color='k',linewidth=.5)

ax.axhline(y=maxthresh,color='k',linewidth=.5,linestyle='--')

ax.axhline(y=minthresh,color='k',linewidth=.5,linestyle='--')

ax.plot(xx,high,marker='o',linestyle='none',color='r')

ax.plot(xx,low,marker='o',linestyle='none',color='b')

#plt.fill_between(xx,y1=maxthresh,y2=diff,where=diff>=maxthresh,color='r')

#plt.fill_between(xx,y1=minthresh,y2=diff,where=diff<=minthresh,color='b')

ax.set_title(field + ' ' + region + ' ' + skey + ' ' + sea)

ax=axs[1]

nnhi,binshi,patcheshi = ax.hist(high.compressed(),color='r',alpha=0.5,normed=True,histtype='stepfilled')

nnlo,binslo,patcheslo = ax.hist(low.compressed(),color='b',alpha=0.5,normed=True,histtype='stepfilled')

nntot,binstot,patchestot = ax.hist(diff,color='0.5',alpha=0.5,normed=True,histtype='stepfilled')

else:

# Plot pvals for mean anomaly and standard dev

# @@ for some reason if all sims are in dataframe, only some are plotted..??

fig,axs=plt.subplots(1,2)

fig.set_size_inches(12,4)

ax=axs[0]

tpvaldft.plot(linestyle='None',color=ccm.get_colordict().values(),marker='s',markersize=6,ax=ax)

ax.axhline(y=siglevel,color='k')

ax.set_ylabel('PVAL of T statistic for mean')

ax.set_title(fdict['fieldstr'] + ' ' + region)

ax.set_ylim((0,1))

ax=axs[1]

fpvaldft.plot(linestyle='None',colors=ccm.get_colordict().values(),marker='s',markersize=6,ax=ax)#make legend better@@

ax.axhline(y=0.05,color='k')

ax.set_ylabel('PVAL of F statistic for variance')

ax.set_title(fdict['fieldstr'] + ' ' + region)

ax.set_ylim((0,1))

if printtofile:

fig.savefig(fdict['fieldstr']+'_ftpvals' + savestr + '_' + region + '.pdf')

if plottype=='plotscatter':

print 'Implement plotscatter! @@@'

## # get datablobs of two fields

## dblob = sfnc.calc_seasons(fdict,coords,sims,seas=seasons,

## loctimesel=timesel,info=infodict,calctype='regmeanwithtime',

## effdof=effdof)

## infodict['region'] = region2

## dblob2 = sfnc.calc_seasons(fdict,coords,sims,seas=seasons,

## loctimesel=timesel,info=infodict,calctype='regmeanwithtime',

## effdof=effdof)

## sfnc.plot_scatter_regions()

if testhadisst:

print '@@testhadisst not implemented'