#!/usr/bin/env python
# Wenchang Yang (wenchang@princeton.edu)
# Tue Jun 29 16:14:32 EDT 2021
if __name__ == '__main__':
    from misc.timer import Timer
    tt = Timer(f'start {__file__}')
import sys, os.path, os, glob, datetime
import xarray as xr, numpy as np, pandas as pd, matplotlib.pyplot as plt
#more imports
import geoxarray
#from misc.landmask import flagland
#import salem
from misc import get_kws_from_argv
#
if __name__ == '__main__':
    tt.check('end import')
#
#start from here
model = os.path.basename(os.getcwd())
#shapefile = '../state_Rio_Grande_do_Sul/43UF2500G.shp'
#shdf = salem.read_shapefile(shapefile)
#region = 'Rio Grande'
region = get_kws_from_argv('region', 'southern')
#ifile = 'precip_FLOR_HistRCP45_tigercpu_intelmpi_18_576PE_10ens_1860-2100_atmos_daily.landfldmean.20_28.5_36_42.5.nc'
#ifile = glob.glob(f'precip_*.landfldmean.20_28.5_36_42.5.nc')[0]
if region == 'southern':
    ifile = glob.glob(f'precip_*_{region}_masked.nc')[0]
elif region == 'northern':
    ifiles = glob.glob('precip_*_hurricaneHelene.nc')
    ifiles.sort()
#if model == 'AM2.5C360':
#    ifile = 'precip_AM2.5C360_amipHadISSTrcp45_tigercpu_intelmpi_18_1080PE_3ens_1871-2100_atmos_daily.region.300_312_-35_-25.RioGrande.nc'
#elif modle = 'FLOR':
#    ifile = 'precip_FLOR_HistRCP45_tigercpu_intelmpi_18_576PE_10ens_1860-2100_atmos_daily.region.300_312_-35_-25.RioGrande.nc'
dsname =  'precip'
daname = dsname
units = 'mm/day'
#years = slice('1981', '2010')
#years = slice('1990', '2020')
years = slice('1991', '2020')
#title = f'AM2p5C360 AMIP 3ens East Africa {years.start}-{years.stop}'
title = f'{daname} {model} {region}  {years.start}-{years.stop}'
tag = '__' + title.replace(' ', '_')
ofile_nc = __file__.replace('.py', f'{tag}.nc')

if os.path.exists(ofile_nc):
    ds = xr.open_dataset(ofile_nc)
    print('[loaded]:', ofile_nc)
else:
    print('loading...')
    if region == 'southern':
        da = xr.open_dataarray(ifile).sel(time=years).load()
    elif region == 'northern':
        da = xr.open_mfdataset(ifiles)[daname].sel(time=years).load()
    print('calculating...')
    #da = da.where(landflag(da)>0.5).geo.fldmean()
    #da = da.geo.fldmean()
    #da = da.assign_coords(lon=da.lon.values-360).salem.roi(shape=shdf).geo.fldmean()
    da = da.geo.fldmean()
    damean = da.groupby('time.dayofyear').mean(['time', 'ens'])
    da_quantile = da.groupby('time.dayofyear').quantile([0.025, 0.17, 0.83, 0.975], dim=['time', 'ens'])
    ds = xr.Dataset(dict(m=damean, q=da_quantile))
    print('saving...')
    ds.to_netcdf(ofile_nc)
    print('[saved]:', ofile_nc)
 
if __name__ == '__main__':
    from wyconfig import * #my plot settings
    #ds = ds.pipe(lambda x: x-273.15)
    fig, ax = plt.subplots()
    #rotate dayofyear
    roll_time = False
    if roll_time:
        ndays = 31+29+31+30+31+30 #Jan-Jun days
        ds = ds.roll(dayofyear=-ndays, roll_coords=True)
        ds = ds.assign_coords(dayofyear=ds.dayofyear.where(ds.dayofyear>ndays, other=ds.dayofyear+366))
    else:
        ndays = 0
    ds.m.plot(color='k', lw=2, label='mean')

    for ii,qi in enumerate(ds['quantile'].values):
        ds.q.sel(quantile=qi).plot(ax=ax, color=f'C{ii}', label=f'{qi*100:.1f}%', lw=1)
    xticks = np.cumsum([1, 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30])
    xticklabels = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
    #roll xticks
    if roll_time:
        xticks = np.roll(xticks, -6)
        xticks = np.where(xticks>ndays, xticks, xticks+366)
        xticklabels = np.roll(xticklabels, -6)
    ax.set_xticks(xticks)
    ax.set_xticklabels(xticklabels, ha='left')

    ax.set_xlim(1+ndays, 366+ndays)
    ax.set_xlabel('')
    ax.set_ylabel(f'{daname} [{units}]')
    ax.set_title('')
    #ax.text(0,1, f' {360-lons.start}-{360-lons.stop}W, {lats.start}-{lats.stop}N, land', transform=ax.transAxes,
    #    ha='left', va='top')
    #ax.legend(loc='upper left')
    ax.legend()
    ax.set_title(title, loc='left')
    
    #savefig
    if 'savefig' in sys.argv or 's' in sys.argv:
        figname = __file__.replace('.py', f'{tag}.png')
        wysavefig(figname)
    tt.check(f'**Done**')
    plt.show()