NOW code for Tasman Sea boxplot

TasmanSeaBoxPlot.py

if paper_case=='20170113_EACnowClimateChange01':
    section_description_dict=collections.OrderedDict()
    section_description_dict['AB']=[ 60, 61,217,226,0,54,'zonal'            ]
    section_description_dict['BC']=[ 60, 61,226,303,0,54,'zonal'            ]
    section_description_dict['GD']=[ 85,128,311,312,0,54,'meridional'       ]
    section_description_dict['FE']=[127,128,239,247,0,54,'zonal'            ]
    section_description_dict['ED']=[127,128,247,311,0,54,'zonal'            ]
    section_description_dict['IH']=[ 46,58 ,212,213,0,54,'meridional'       ]
    section_description_dict['JK']=[ 85,86 ,224,231,0,54,'zonal'            ]
    section_description_dict['PQ']=[ 68,80 ,212,213,0,54,'meridional'       ]
    section_description_dict['RS']=[ 58,66 ,306,307,0,54,'meridional'       ]
    section_description_dict['DU']=[128,148,311,312,0,54,'meridional'       ]
    section_description_dict['VW']=[147,148,230,295,0,54,'zonal'            ]
    section_description_dict['XU']=[147,148,297,312,0,54,'zonal'            ]
    
def main_pretty_new_sections_now_climatechange(hdf_in,evaluation=False,ctl_anom_only=''):
    """main function to actually create the plot, re-factored to do bar charts with seaborn...
    :hdf_in: HDFSTORE from ./mk_nemo_UV_transport.py
    :evaluation(optional): plots the evaulation numbers
    :ctl_anom_only(optional): remove the CTRL simulation
    :returns: @todo
    """

    import seaborn as sns #bit dodgy but don't want to mess up my other plots
    sns.set_style("whitegrid", {'axes.grid' : False})

    df=pd.HDFStore(hdf_in).DF

    df['RG94']=np.nan
    df['CARS']=np.nan
    df['OFAM']=np.nan
    df['MOM']=np.nan
    df.loc['ED','RG94']=17.7
    df.loc['GD','RG94']=3.6
    df.loc['FE','RG94']=27.4
    df.loc['BC','RG94']=2.0
    df.loc['IH','RG94']=8.1
    df.loc['AB','RG94']=8.1

    df.loc['ED','CARS']=14.8
    df.loc['GD','CARS']=3.7 
    df.loc['FE','CARS']=25.8
    df.loc['BC','CARS']=3.2 
    df.loc['IH','CARS']=10.8
    df.loc['AB','CARS']=9.7 
                      
    df.loc['ED','OFAM']=8.5 
    df.loc['GD','OFAM']=12.6
    df.loc['FE','OFAM']=27.5
    df.loc['BC','OFAM']=2.7 
    df.loc['IH','OFAM']=10.8
    df.loc['AB','OFAM']=9.7 

    df.loc['ED','MOM']=7.9 
    df.loc['GD','MOM']=9.0 
    df.loc['FE','MOM']=20.4
    df.loc['BC','MOM']=5.6 
    df.loc['IH','MOM']=np.nan
    df.loc['AB','MOM']=9.7 

    # df=pd.DataFrame(\
    # {'RG94':[17.7,3.6,27.4,2.0,8.1,8.1],\
    # 'CARS':[14.8,3.7 ,25.8,3.2 ,10.8,9.7 ],\
    # 'OFAM':[8.5 ,12.6,27.5,2.7 ,10.8,9.7 ],\
    # 'MOM':[7.9 ,9.0 ,20.4,5.6 ,np.nan,9.7 ],\
    # },index=['ED','DG','FE','BC','IH','AB'])

    plt.close('all')
    #width then height
    #fig=plt.figure()
    fig=plt.figure(figsize=(10.0,10.0))
    ax=fig.add_subplot(1, 1,1)

    #plotmask(ax,filled=True)
    plotmask(ax,filled=True)

    if evaluation:
        plotpointslines(ax,jk=False,pts=points)
    else:
        plotpointslines(ax,jk=True,pts=points)

    # sm.pl_inset_title_box(ax,'b',bwidth="5%",location=2)

    #plotmask(ax,strechgrid=False,filled=True)
    #ax.set_title('msg')
    ax.set_xlabel('Longitude')
    ax.set_ylabel('Latitude')

    colour_dict=collections.OrderedDict()
    if evaluation:
        colour_dict['RG94']='#e6ab02'
        colour_dict['CARS']='#a6761d'
        colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP']] ='#e41a1c'

        #force the order we like...
        order=\
        [\
        'RG94',\
        'CARS',\
        'now_cordex24_BMJv2_BILAP'\
        ]
    else:

        if ctl_anom_only=='':
            colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP']]        =sm.exp_colours_now()[0]['now_cordex24_BMJv2_BILAP']
        colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP_CO2']]    =sm.exp_colours_now()[0]['now_cordex24_BMJv2_BILAP_CO2']
        colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP_OCEBDY']] =sm.exp_colours_now()[0]['now_cordex24_BMJv2_BILAP_OCEBDY']
        colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP_ATMBDY']] =sm.exp_colours_now()[0]['now_cordex24_BMJv2_BILAP_ATMBDY']
        colour_dict[sm.exp_names_now()['now_cordex24_BMJv2_BILAP_ALL']]    =sm.exp_colours_now()[0]['now_cordex24_BMJv2_BILAP_ALL']

        #force the order we like...
        order=\
        [\
        'now_cordex24_BMJv2_BILAP',\
        'now_cordex24_BMJv2_BILAP_CO2',\
        'now_cordex24_BMJv2_BILAP_OCEBDY',\
        'now_cordex24_BMJv2_BILAP_ATMBDY',\
        'now_cordex24_BMJv2_BILAP_ALL'\
        ]
        sm.pl_inset_title_box(ax,'d',bwidth="5%",location=2)

    # colour_dict['OFAM']='#fb9a99'
    # colour_dict['MOM']= '#fccde5'

    #when we remove the mean we no longer want to plot it..
    if ctl_anom_only!='' and not evaluation:
        order.remove(ctl_anom_only)

    ins_custom_legend(ax,colour_dict,location=4,ncols=1)
    if evaluation:
        secs=['ED','DG','FE','AB','BC','IH']
    else:
        secs=['ED','DG','FE','AB','BC','JK','IH']

    for section_name in secs:
    # for section_name in ['ED','DG','FE','AB','BC']:

        x=np.average([points[section_name[0:1]][0],points[section_name[1:]][0]])
        x=points[section_name[0:1]][0]
        y=points[section_name[0:1]][1]
        # print section_name
        if section_name=='ED':
            x=x+10
            y=y-1.3
        elif section_name=='DG':
            x=x-.3
            y=y-8.6
        elif section_name=='FE':
            x=x+3.2
            y=y-6
        elif section_name=='IH':
            x=x-0.1
            y=y-2.5
            # print section_name,x,y
        elif section_name=='AB':
            x=x+5
            y=y-3.3
            # print section_name,x,y
        elif section_name=='BC':
            x=x+12
            y=y+1.2
        elif section_name=='JK':
            x=x+4.4
            y=y-3.8

        # if section_name=='HI':
            # section_name='IH'

        if section_name=='DG':
            section_name='GD'

        x,y=axis_data_coords_sys_transform(ax,x,y,True)
        #last two numbers indicate size... (width then height)
        inset_axes =  fig.add_axes([x,y,0.17,0.1])
        data=df.ix[section_name]


        #check for negatives
        if data.iloc[0]<0:
            data=np.abs(data)
            # data.sort(ascending=False,inplace=True)
        # else:
        # data.sort(ascending=True,inplace=True)

        #remove the CTRL simulation from the transports
        if ctl_anom_only!='' and not evaluation:
            data=data-data[ctl_anom_only]
            # print order

        b=sns.barplot(order,data.ix[order],palette=colour_dict.values(),ax=inset_axes)

        #turn axis stuff off
        inset_axes.axes.get_yaxis().set_visible(False)
        inset_axes.axes.get_xaxis().set_visible(False)
        b.set_yticks([])
        sns.despine(ax=b, left=True, bottom=True)

        #put title up high
        plt.text(0.5, 1.28, section_name,
                 horizontalalignment='center',
                  fontsize=13,
                   transform = inset_axes.transAxes)
        # inset_axes.set_title(section_name)
        # inset_axes.set_xlabel(section_name)

        #insert size of transport..
        for p in b.patches:
            b.annotate(
            s='{:.1f}'.format(p.get_height()),
            xy=(p.get_x()+p.get_width()/2.,p.get_height()),
            ha='center',va='center',
            xytext=(0,10),
            textcoords='offset points'
            )

    make_arrow(ax,'meri','ED','up',points)
    make_arrow(ax,'meri','FE','down',points)
    make_arrow(ax,'meri','BC','up',points)
    make_arrow(ax,'meri','AB','down',points)
    make_arrow(ax,'zonal','DG','right',points)
    make_arrow(ax,'zonal','HI','left',points)
    if not evaluation:
        make_arrow(ax,'meri','JK','down',points)

    #since it's a schematic AK and I have decided to kill axis labels...
    ax.set_xticklabels([])
    ax.set_yticklabels([])

    # sm.ins_legend(ax,exp_subset=None,lw=3,location=1,invisible=False,ncols=1)
    # sm.change_tick_labels_add_dirs(ax)

    # plt.show()
    if evaluation:
        path_to_png=output_folder+'nemo_UV_transports_'+'pretty_bar_charts_fxdtrans_evaluation.png'
    else:
        path_to_png=output_folder+'nemo_UV_transports_'+'pretty_bar_charts_fxdtrans.png'

    if ctl_anom_only!='' and not evaluation:
        path_to_png=output_folder+'nemo_UV_transports_'+'pretty_bar_charts_fxdtrans_anomalies.png'

    plt.savefig(path_to_png,dpi=300)
    #plt.savefig(output_folder+'nemo_UV_transports_'+exptwoname+'.pdf')
    it.AddTrkr(path_to_png,{'Created with':os.path.realpath(__file__),'time':datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),'machine':socket.gethostname()})
    # print path_to_png
    lg.info("Plot path is: " + path_to_png)
    #plt.show()
    return