library(plyr)
library(dplyr)
library(ggdistribute)
library(ggplot2)
library(ggExtra)
library(scales)
library(gridExtra)
library(stringr);
library(DBI);

priceg = ".*"
mold = "XTG154"
colgrp = ".*"
branding = ".*"
outlier = 300
xfact = 5
yfact = 5
xtrans = "identity"
ytrans = "identity"
lprice = .01
uprice = 30.00
pqty = 2000


#Ipng(filename="target.png")

#prod_plot <- function(priceg, mold, colgrp, branding, outlier, xfact, yfact,xtrans, ytrans, lprice, uprice, pqty ) {
    sql = paste("SELECT * FROM rlarp.rlang_plot('",mold,"','",priceg,"','",colgrp,"','",branding,"',",lprice,",",uprice,") x",sep="");

    con <- dbConnect(RPostgres::Postgres(),dbname = 'ubm', 
                     host = 'usmidlnx01',
                     port = 5030,
                     user = 'report',
                     password = 'report')

    d <- dbGetQuery(con, sql)

    dbDisconnect(con)

    #-----each graph is composed of 2 pieces when doing the facet() pivot, these 2 pieces make up the plot defition-----
    d$f7 <- substring(d$mold,1,7)
    #d$dim1 <- trimws(paste(d$f7,d$v1ds));
    #d$dim1 <- trimws(paste(d$f7,d$colgrp,d$brnd,d$package,d$suffix,d$kit));
    d$dim1 <- trimws(paste(d$base_item,d$colgrp,d$brnd));
    d$dim2 <- trimws(paste(d$chgrp));
    d$plot <- trimws(paste(d$dim1,d$dim2));
    d$sub <- trimws(paste("v1:",d$coltier,".",substring(d$brnd,1,1),".",d$package,".",d$suffix,".",d$kit));
    #d$sub <- trimws(paste(d$oseas));
    #d$sub <- trimws(paste(d$geo));
    d$qty = d$qty/1000;
    #-----need to include credits------
    d$volmin = 0.0001;
    d$season = factor(d$oseas);
    #-----build widths for how many scenarios are present----------------------------------------------------------------
    dim1 <- data.frame(unique(d$dim1));
    var.dim1 = nrow(dim1);
    dim2 <- data.frame(unique(d$dim2));
    var.dim2 = nrow(dim2);
    #-----------need to do an aggregate to consolidate to single customer point
    d <- subset(d,chgrp != "X", promo != "Excess and Obsolete");
    #-------------------------eliminate outliers-------------------------------------------------------------------------
    dx <- boxplot.stats(d$price, coef = outlier);
    ex <- data.frame(dx$out);
    #ex;  #list the excluded outlier prices
    colnames(ex)[1] = "price";
    outl <- inner_join(d,ex, by = "price");
    outl;
    d <- anti_join(d,ex, by = "price");
    #---------switch to log axis if there are still outliers with a coefficient 3----------------------------------------
    var.trans = "identity"
    if (nrow(data.frame(boxplot.stats(d$price, coef = 3)$out)) >= 1){
        var.trans = "log2"
    };
    

    glob <- ddply(d, .(), summarise,
            Volume=round(sum(qty),0),
            Sales=round(sum(sales),0),
            WeightedAvg=round(sum(sales)/sum(qty),4),
            Mean=round(mean(price),4),
            StdDev=round(sd(price),4),
            Target=round(mean(target_price),4),
            AnyMax=round(max(c(price,target_price)),4),
            AnyMin=round(min(c(price,target_price)),4),
            PriceMin = round(min(price),4),
            PriceMax = round(max(price),4),
            VolumeMin  = round(min(pmax(qty,volmin)),4),
            VolumeMax = round(max(qty),1),
            VolumeSD=round(sd(pmax(qty,volmin)),4)
    );
    #targets <- ddply(d, .(dim1, dim2, plot,mold,chan,colgrp, brnd), summarise,
    targets <- ddply(d, .(dim2, v1ds, dim1, plot,mold,chan,colgrp, brnd), summarise,
            Volume=round(sum(qty),0),
            Sales=round(sum(sales),0),
            WeightedAvg=round(sum(sales)/sum(qty*1000),4),
            Mean=round(mean(price),4),
            StdDev=round(sd(price),4),
            Target=round(mean(target_price),4),
            HexCol = min(hex)
    );
    seas <- ddply(d, .(dim1, dim2, plot, oseas), summarise,
            Volume=round(sum(qty),0),
            Sales=round(sum(sales),0),
            WeightedAvg=round(sum(sales)/sum(qty*1000),4),
            Mean=round(mean(price),4),
            StdDev=round(sd(price),4),
            Target=round(mean(target_price),4)
    );
    #-----------------blank dataframe in case there is no data for a scenario-----------------
    blank <- glob
    blank$customer = 'NO DATA'
    blank$oseas = 2020
    blank$season = '2020'
    blank$qty = blank$VolumeSD
    blank$price = blank$Mean
    #blank;

    yr1 <- subset(seas, oseas == 2020);
    yr2 <- subset(seas, oseas == 2021);
    dir_t <- subset(targets, chan == "DIR");
    drp_t <- subset(targets, chan == "DRP");
    whs_t <- subset(targets, chan == "WHS");

    anno <- data.frame(unique(d[c("plot","dim2","dim1","mold","colgrp","brnd")]));
    anno <- data.frame(anno,qty=c(Inf),price=c(Inf),hjustvar = c(1),vjustvar = c(1));

    anno <- merge(x = anno, y = yr1[ , c("plot","Mean","WeightedAvg", "StdDev","Volume")], by = "plot", all.x=TRUE);
    names(anno)[names(anno)=="Mean"] <- "yr1_mn";
    names(anno)[names(anno)=="WeightedAvg"] <- "yr1_wa";
    names(anno)[names(anno)=="StdDev"] <- "yr1_sd";
    names(anno)[names(anno)=="Volume"] <- "yr1_vo";

    anno <- merge(x = anno, y = yr2[ , c("plot","Mean","WeightedAvg", "StdDev","Volume")], by = "plot", all.x=TRUE);
    names(anno)[names(anno)=="Mean"] <- "yr2_mn";
    names(anno)[names(anno)=="WeightedAvg"] <- "yr2_wa";
    names(anno)[names(anno)=="StdDev"] <- "yr2_sd";
    names(anno)[names(anno)=="Volume"] <- "yr2_vo";

    anno <- merge(x = anno, y = dir_t[ , c("plot","Target")], by = "plot", all.x=TRUE);
    names(anno)[names(anno)=="Target"] <- "t_dir";

    anno <- merge(x = anno, y = drp_t[ , c("plot","Target")], by = "plot", all.x=TRUE);
    names(anno)[names(anno)=="Target"] <- "t_drp";

    anno <- merge(x = anno, y = whs_t[ , c("plot","Target")], by = "plot", all.x=TRUE);
    names(anno)[names(anno)=="Target"] <- "t_whs";

    csv <- anno;
    csv <- subset(csv, select = c(mold, dim2, colgrp, brnd, yr1_mn, yr2_mn, yr1_wa, yr2_wa, t_dir, t_drp, t_whs));
    csv$t_dir_rev = csv$t_dir;
    csv$t_drp_rev = csv$t_drp;
    csv$t_whs_rev = csv$t_whs;
    names(csv)[names(csv)=="dim2"] <- "chgrp";
    csv;
    #write.csv(csv, file = paste("//home/ptrowbridge/pt_share/",file_name,"_TRG.csv",sep=""), row.names = FALSE);

    p=ggplot(d, aes(x=qty, y=price, color=v1ds)) +
        #scale_color_manual(values=c("#F44336", "#E91E63", "#9C27B0","#673ab7","#3f51b5","#2196f3","#03a9f4","#00bcd4","#009688","#4caf50","#8bc34a","#8bc34a","#ffeb3b","#ffc107")) +
        geom_point(size=2) +
        geom_text(data = anno,
            aes(
                x=qty,y=price,
                color = NULL,
                hjust=hjustvar,vjust=vjustvar,
                label=paste(
                    " mean  |  wavg  | stdd   |  vol    \n",
                    "-------|--------|--------|---------\n",
                    "PY(black): ",
                    #----------mean-------------------------------
                    str_pad(
                        format(round(yr1_mn, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "|",
                    #----------weighted average-------------------
                    str_pad(
                        format(round(yr1_wa, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "
                    ),
                    #----------standard deviation-----------------
                    "|",
                    str_pad(
                        format(round(yr1_sd, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "|",
                    #----------volume-----------------------------
                    str_pad(
                        format(round(yr1_vo/1000, 4), nsmall = 4,width = 7),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "\n",
                    "CY(green): ",
                    #----------mean-------------------------------
                    str_pad(
                        format(round(yr2_mn, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "|",
                    #----------weighted average-------------------
                    str_pad(
                        format(round(yr2_wa, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "
                    ),
                    #----------standard deviation-----------------
                    "|",
                    str_pad(
                        format(round(yr2_sd, 4), nsmall = 4),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "|",
                    #----------volume-----------------------------
                    str_pad(
                        format(round(yr2_vo/1000, 4), nsmall = 4,width = 7),
                        width = 6,
                        side = "both",
                        pad = " "),
                    "\n",
                    #format(round(yr2_mn, 4), nsmall = 4),"|",format(round(yr2_wa, 4), nsmall = 4),"|",format(round(yr2_sd, 4), nsmall = 4),"|",format(round(yr2_vo/1000, 4), nsmall = 4,width = 7),"\n",
                    "                                   \n",
                    "  dir (b)  |  drp (y)  |  whs (r)  \n",
                    "-----------|-----------|-----------\n",
                    "Targets: ",
                    str_pad(
                        format(round(t_dir, 4), nsmall = 4),
                        width = 9, 
                        side ="both",
                        pad=" "),
                    "|",
                    str_pad(
                        format(round(t_drp, 4), nsmall = 4),
                        width = 9,
                        side = "both",
                        pad = " "),
                    "|",
                    str_pad(
                        format(round(coalesce(t_whs,0), 4), nsmall = 4),
                        width = 10,
                        side = "both",
                        pad = " ")
                )
            ),
            family="Courier",
            size = 3,
            #use check_overlap to avoid doubling up the price info print, it will print over top of itself based on the color=sub count of uniques
            check_overlap=TRUE
        ) +
        geom_text(aes(label=customer),size=3, vjust = 2, hjust = 0, check_overlap=TRUE) +
        facet_grid(dim2~dim1) +
        #facet_grid(chgrp~plot) +
        #facet_wrap(plot) +
        geom_hline(data=yr1, aes(yintercept=Mean),linetype="dashed", size=.5, colour="black") +
        #geom_hline(data=yr1, aes(yintercept=Mean - StdDev),linetype="dashed", size=.5, colour="black") +
        #geom_hline(data=yr1, aes(yintercept=Mean - StdDev * 2),linetype="dashed", size=.5, colour="black") +
        geom_hline(data=yr1, aes(yintercept=WeightedAvg),linetype="solid", size=.5, colour="black") + 
        geom_vline(aes(xintercept = pqty/1000)    ,linetype = "dashed",size = .5, colour = "orange") +
        #geom_vline(aes(xintercept = pqty/1000*8)  ,linetype = "dashed",size = .5, colour = "grey") +
        geom_vline(aes(xintercept = pqty/1000*8) ,linetype = "dashed",size = .5, colour = "grey") +
        #geom_hline(data=yr2, aes(yintercept=Mean),linetype="dashed", size=.5, colour="green") +
        #geom_hline(data=yr2, aes(yintercept=Mean - StdDev),linetype="dashed", size=.5, colour="green") +
        #geom_hline(data=yr2, aes(yintercept=Mean - StdDev * 2),linetype="dashed", size=.5, colour="green") +
        #geom_hline(data=yr2, aes(yintercept=WeightedAvg),linetype="solid", size=.5, colour="green") +
        geom_hline(data=drp_t, aes(yintercept=Target, color=v1ds),linetype="solid", size=.5) +
        geom_hline(data=dir_t, aes(yintercept=Target, color=v1ds),linetype="solid", size=.5) +
        geom_hline(data=whs_t, aes(yintercept=Target, color=v1ds),linetype="solid", size=.5) +
        #scale_y_continuous(breaks=seq(0, 10, round(glob$StdDev * .5,2))) + 
        scale_y_continuous(
            #breaks=seq(glob$PriceMin, glob$PriceMax, round(glob$StdDev * .5,4)),
            breaks = pretty_breaks(n=20),
            limits = c(glob$AnyMin, glob$AnyMax), trans = ytrans
        ) +
        scale_x_continuous(
            #breaks=seq(glob$VolumeMin, glob$VolumeMax, round(glob$VolumeSD * 1.0,4)),
            #breaks = pretty_breaks(n=10),
            limits = c(glob$VolumeMin, glob$VolumeMax*1.1), trans = xtrans
        ) +
        #scale_x_continuous(trans='log2') +
        #scale_x_continuous(breaks=seq(0,1000,round(glob$VolumeSD * 1,2)), trans = 'log2') +
        #geom_label(colour = "white", fontface = "bold") +
        #geom_text(aes(label=ds$ship_group),position = position_dodge(width=.9),  size=2) + 
        theme(legend.position="none");
    cp_pvt = p + theme_bw();
    #targets;
    options(
        repr.plot.width=var.dim1*xfact, 
        repr.plot.height=var.dim2*yfact
    );
    cp_pvt;
    ggsave("targetggsave.png")
    #dev.print(png, 'targetpl.png')

    #dev.off()
#};