---
title: "`r paste(BASE_SCENARIO_NAME, 'vs.', BUILD_SCENARIO_NAME, 'Calibration Summary')`"
output:
  flexdashboard::flex_dashboard:
    orientation: columns
    vertical_layout: fill
    theme: spacelab
    social: menu
    source_code: embed
runtime: shiny
---


```{r Setup}
opts_knit$set(root.dir = SYSTEM_APP_PATH)
knit_hooks$set(optipng = hook_optipng)

TILE_PROVIDER = providers$Esri.WorldImagery
```

```{r setpar}
knitr::opts_knit$set(global.par = TRUE)
```


```{r ggplot_Theme}
theme_db <- theme_bw() + theme(plot.margin = unit(c(10,10,20,10),"pt"))
```

```{r Helper_Functions}
compare_bar_plotter <- function(base, build, base_name, build_name, xvar, yvar,
                        xlabel = xvar, ylabel = yvar, position = "dodge",
                        xrotate = FALSE, yrotate = FALSE, coord_flip = FALSE,
                        title = "", left_offset = 0, bottom_offset = 0){

  base$grp <- base_name
  build$grp <- build_name
  colnames(build) <- colnames(base)

  df <- rbind(base, build)

  p <- ggplot(df, aes_string(x = xvar, y = yvar)) +
    geom_bar(stat = "identity", aes(fill = grp), position = position) +
    xlab(xlabel) + ylab(ylabel) +
    labs(fill = "") +
    ggtitle(title) +
    theme(axis.text.x=element_text(angle=50, size=1, vjust=0.5)) +
    theme(axis.text.y=element_text(angle=50, size=1, vjust=0.5)) +
    theme_bw()

  if (xrotate) {
    p <- p + theme(axis.text.x = element_text(angle = 45, hjust = 1))
  }
  if (yrotate) {
    p <- p + theme(axis.text.y = element_text(angle = 45, hjust = 1))
  }
  if (coord_flip) {
    p <- p + coord_flip()
  }


  p <- plotly_build(p)
  p$layout$margin$l <- p$layout$margin$l+left_offset
  p$layout$margin$b <- p$layout$margin$b+bottom_offset
  return(p)

}

# This function combines two dataframes and returns a data frame with standard field names
# The field names in the two dataframes should be same and should be same as the variable
# names passed to the function
# input parameter - dataframe1, dataframe2, x variable, list of y variables
# renames x and y variables in standard form - xvar, (yvar1, yvar2),...
# Y variables are named in pairs - (yvar1, yvar2), (yvar3, yvar4), ....
# yvar1, yvar3, .. correspond to first dataframe and yvar2, yvar4, .. correspond to second dataframe
# computes proportions for each  y variable variable
get_standardDF <- function(data_df1, data_df2, x, y, grp = "", shared = F){

  #data_df1=base_df
  #data_df2=build_df
  #x="id"
  #y = c("freq_out", "freq_inb")
  #grp = "purpose"
  #shared = T
  #
  # create ID variable to join base and build data
  if(!shared){
    ev1 <- paste("data_df1$id_var <- data_df1$", x, sep = "")
    ev2 <- paste("data_df2$id_var <- data_df2$", x, sep = "")
    eval(parse(text = ev1))
    eval(parse(text = ev2))
  }else{
    if(grp==""){
      stop("group variable not specified")
    }else{
      ev1 <- paste("data_df1$id_var <- paste(data_df1$", grp, ", data_df1$", x, ', sep = "")', sep = "")
      ev2 <- paste("data_df2$id_var <- paste(data_df2$", grp, ", data_df2$", x, ', sep = "")', sep = "")
      eval(parse(text = ev1))
      eval(parse(text = ev2))
    }
  }

  data_df <- data_df1

  # rename variables to standard names
  names(data_df)[names(data_df) == x] <- 'xvar'
  if(shared){
    if(grp==""){
      stop("group variable not specified")
    }else{
      names(data_df)[names(data_df) == grp] <- 'grp_var'
    }
  }

  for(i in seq(from=1, to=length(y))){
    start_pos <- i*2-1
    yvar1 <- paste('yvar', start_pos, sep = "")
    yvar2 <- paste('yvar', start_pos+1, sep = "")
    names(data_df)[names(data_df) == y[[i]]] <- paste('yvar', start_pos, sep = "")
    eval_expr <- paste("data_df$", yvar2, " <- data_df2$", y[[i]], "[match(data_df$id_var, data_df2$id_var)]", sep = "")
    eval(parse(text = eval_expr))
  }
  
  # data_df[is.na(data_df)] <- 0

  #data_df$grp_var <- as.character(data_df$grp_var)

  # compute proportions for y variables
  for(i in seq(from=1, to=length(y))){
    start_pos <- i*2-1
    prop_var1 <- paste('prop', start_pos, sep = "")
    y_var1    <- paste('yvar', start_pos, sep = "")
    prop_var2 <- paste('prop', start_pos+1, sep = "")
    y_var2    <- paste('yvar', start_pos+1, sep = "")
    if(shared){
      if(grp==""){
        stop("group variable not specified")
      }else{
        eval_expr1 <- paste("data_df <- data_df %>% group_by(grp_var) %>% mutate(", prop_var1, " = prop.table(", y_var1, "))", sep = "")
        eval_expr2 <- paste("data_df <- data_df %>% group_by(grp_var) %>% mutate(", prop_var2, " = prop.table(", y_var2, "))", sep = "")
      }
    }else{
      eval_expr1 <- paste("data_df <- data_df %>% mutate(", prop_var1, " = prop.table(", y_var1, "))", sep = "")
      eval_expr2 <- paste("data_df <- data_df %>% mutate(", prop_var2, " = prop.table(", y_var2, "))", sep = "")
    }

    eval(parse(text = eval_expr1))
    eval(parse(text = eval_expr2))
  }

  # set all NAs to zero
  # data_df[is.na(data_df)] <- 0

  if(!shared){
    return(data_df)
  }else{
    data_sd <- SharedData$new(data_df, ~grp_var)
    return(data_sd)
  }
}

# This function returns a SharedData object for creating comparison density plots
# input parameter - dataframe1, dataframe2, x variable, list of y variables,
# grouping variable, names of each run
# The function expects same field names across both dataframes
# renames x and y variables in standard form - xvar, yvar1, yvar2,...
# computes proportions for each  y variable variable for each group and run
# combines two dataframe and adds a run identifier
get_sharedData <- function(data_df1, data_df2, run1_name = 'base', run2_name = 'build',
                           x, y, grp){

  # rename variables to standard names
  names(data_df1)[names(data_df1) == x] <- 'xvar'
  names(data_df1)[names(data_df1) == grp] <- 'grp_var'
  for(i in 1:length(y)){
    names(data_df1)[names(data_df1) == y[[i]]] <- paste('yvar', i, sep = "")
  }

  names(data_df2)[names(data_df2) == x] <- 'xvar'
  names(data_df2)[names(data_df2) == grp] <- 'grp_var'
  for(i in 1:length(y)){
    names(data_df2)[names(data_df2) == y[[i]]] <- paste('yvar', i, sep = "")
  }

  # compute proportions for y variables
  data_df1 <- group_by(data_df1, grp_var)
  for(i in 1:length(y)){
    prop_var <- paste('prop', i, sep = "")
    y_var    <- paste('yvar', i, sep = "")
    eval_expr <- paste("data_df1 <- data_df1 %>% mutate(", prop_var, " = prop.table(", y_var, "))", sep = "")
    eval(parse(text = eval_expr))
  }

  data_df2 <- group_by(data_df2, grp_var)
  for(i in 1:length(y)){
    prop_var <- paste('prop', i, sep = "")
    y_var    <- paste('yvar', i, sep = "")
    eval_expr <- paste("data_df2 <- data_df2 %>% mutate(", prop_var, " = prop.table(", y_var, "))", sep = "")
    eval(parse(text = eval_expr))
  }

  # add run identifiers
  data_df1$run <- run1_name
  data_df2$run <- run2_name

  # combine dataframes
  data_df <- rbind(data_df1, data_df2)

  # set all NAs to zero
  data_df[is.na(data_df)] <- 0

  data_sd <- SharedData$new(data_df, ~grp_var)
  return(data_sd)
}

# This function returns bar plot for a given X-Y data frame
# The function expects the data frame columns to be named as
# xvar, yvar1, yvar2...
# function plots only two series at a time
# which two y series to plot is determined by the index variable
# index==1 :- yvar1, yvar2, index==2 :- yvar,3,4 and so on
# names of series to be plotted should also be passed as a list argument
# number of elements in names list determines the number of series to be added
plotly_bar_plotter <- function(data, type = 'bar', xlabel = "", ylabel = "", percent = FALSE,
                               title = "", height = 0, width = 0, ynames = c(""), left_offset = 0,
                               bottom_offset = 0, tickformat = ".2", hoverformat = "", tickangle = 0, index = 1, tickvals = c(), ticktext = c()){
  #initial setup
  start_pos <- 2*index - 1
  exp_tickvals <- ifelse(length(tickvals)>0, ', tickvals = tickvals', "")
  exp_ticktext <- ifelse(length(ticktext)>0, ', ticktext = ticktext', "")

  #generate plot
  if(!percent){
    ylab <- ifelse(ylabel=="", "Percent", ylabel)
    hformat <- ifelse(hoverformat=="", '.1f', hoverformat)
    eval_expr <- paste("p <- plot_ly(data, x = ~xvar, y = ~yvar", start_pos, ", type = type, name = ynames[[1]]) %>% ",
                       "add_trace(y = ~yvar", start_pos+1, ", name = ynames[[2]]) %>% ",
                       "layout(yaxis = list(hoverformat = hformat, title = ylab, tickformat = tickformat), xaxis = list(title = xlabel, tickangle = tickangle", exp_tickvals, exp_ticktext, "), barmode = 'group')", sep = "")
    eval(parse(text = eval_expr))
  }else{
    ylab <- ifelse(ylabel=="", "Frequency", ylabel)
    hformat <- ifelse(hoverformat=="", '.1%', hoverformat)
    eval_expr <- paste("p <- plot_ly(data, x = ~xvar, y = ~prop", start_pos, ", type = type, name = ynames[[1]]) %>% ",
                       "add_trace(y = ~prop", start_pos+1, ", name = ynames[[2]]) %>% ",
                       "layout(yaxis = list(hoverformat = hformat, title = ylab, tickformat = '.0%'), xaxis = list(title = xlabel, tickangle = tickangle", exp_tickvals, exp_ticktext,"), barmode = 'group')", sep = "")
    eval(parse(text = eval_expr))
  }

  p$x$layout$height <- height
  p$x$layout$width <- width
  p$x$layout$margin$b <- p$x$layout$margin$b + bottom_offset
  p$x$layout$margin$l <- p$x$layout$margin$l + left_offset
  return(p)
}

# This function returns a spline plot with fill for a gievn X-Y dataframe
# The function expects the data frame columns to be named as
# x = ~xvar, y = (~yvar1 or prop1),  (~yvar2 or prop2) adn so on (Frequency or Percent),
# which y to use is determined by index parameter (one, two or three)
# and variable differentiating runs as ~run
# The function currebtly plots only one Y variables for each run
plotly_density_plotter <- function(data_df, index = "one", colors=c("orange", "steelblue"), fill = 'tozeroy',
                                   title = "", xlabel = "", ylabel = "", percent = T, alpha = 0.7, tickvals, ticktext, tickangle = 0,
                                   height=400, left_offset = 0, bottom_offset = 0, shape = 'spline', legend = T){
  ##standardize data frame
  #names(data_df)[names(data_df)==xvar]     <- 'xvar'
  #names(data_df)[names(data_df)==yvar]     <- 'yvar1'
  #names(data_df)[names(data_df)==prop_var] <- 'prop1'
  #names(data_df)[names(data_df)==grp]      <- 'run'

  # prepare plot using standardized dataframe
  if(percent){
    ylab <- ifelse(ylabel=="", "Percent", ylabel)

    p <- switch(index,
                "one" = plot_ly(data=data_df,x = ~xvar, y = ~prop1, colors=c("orange", "steelblue"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab, tickformat = ".0%"), showlegend = legend),
                "two" = plot_ly(data=data_df,x = ~xvar, y = ~prop2, colors=c("orange", "steelblue"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab, tickformat = ".0%"), showlegend = legend),
                "three" = plot_ly(data=data_df,x = ~xvar, y = ~prop3, colors=c("orange", "steelblue"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab, tickformat = ".0%"), showlegend = legend)
                )

  }else{
    ylab <- ifelse(ylabel=="", "Frequency", ylabel)

    p <- switch(index,
                "one" = plot_ly(data=data_df,x = ~xvar, y = ~yvar1, colors=c("steelblue", "orange"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab), showlegend = legend),
                "two" = plot_ly(data=data_df,x = ~xvar, y = ~yvar2, colors=c("steelblue", "orange"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab), showlegend = legend),
                "three" = plot_ly(data=data_df,x = ~xvar, y = ~yvar3, colors=c("steelblue", "orange"), color = ~run, fill=fill) %>%
                  add_lines(name=~run,alpha=alpha, line = list(shape = shape)) %>%
                  layout(title = "",xaxis = list(title=xlabel, tickvals = tickvals, ticktext = ticktext, tickangle = tickangle), yaxis = list(title=ylab), showlegend = legend)
                )

    #p <- plot_ly(data=data_df,x = ~xvar, y = ~yvar1, colors=c("steelblue", "orange"), color = ~run, height=400, fill=fill) %>%
    #add_lines(name="",alpha=alpha, line = list(shape = shape)) %>%
    #layout(title = "",xaxis = list(title=xlabel), yaxis = list(title=ylab))
  }

  p$x$layout$height <- height
  p$x$layout$margin$b <- p$x$layout$margin$b + bottom_offset
  p$x$layout$margin$l <- p$x$layout$margin$l + left_offset
  return(p)
}

# This function returns a pie chart
# Input is a 2 column data frame with a label variable and a value variable
plotly_pie_chart <- function(data, label_var, value_var, title = "",
                               height = 0, width = 0, left_offset = 0,bottom_offset = 0, top_offset = 0, shared = F){

  colors <- c('rgb(211,94,96)', 'rgb(128,133,133)', 'rgb(144,103,167)', 'rgb(171,104,87)', 'rgb(114,147,203)')

  if(!shared){
    names(data)[names(data)==label_var] <- 'label_var'
    names(data)[names(data)==value_var] <- 'value_var'

    p <- plot_ly(data, labels = ~label_var, values = ~value_var, type = 'pie',
          textposition = 'outside',
          textinfo = 'label+percent',
          insidetextfont = list(color = '#FFFFFF'),
          marker = list(colors = colors,
                        line = list(color = '#FFFFFF', width = 2)),
          showlegend = FALSE,
          sort = FALSE) %>%
    layout(title = title,
           xaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE),
           yaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE))
  }else{
    eval_expr <- paste("p <- plot_ly(data, labels = ~", label_var, ", values = ~", value_var, ", type = 'pie',
          textposition = 'outside',
          textinfo = 'label+percent',
          insidetextfont = list(color = '#FFFFFF'),
          marker = list(colors = colors,
                        line = list(color = '#FFFFFF', width = 2)),
          showlegend = FALSE,
          sort = FALSE) %>%
    layout(title = title,
           xaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE),
           yaxis = list(showgrid = FALSE, zeroline = FALSE, showticklabels = FALSE))", sep = "")

    eval(parse(text = eval_expr))
  }


  p$x$layout$height <- height
  p$x$layout$width <- width
  p$x$layout$margin$b <- p$x$layout$margin$b + bottom_offset
  p$x$layout$margin$l <- p$x$layout$margin$l + left_offset
  p$x$layout$margin$t <- p$x$layout$margin$t + top_offset
  return(p)
}

lm_eqn <- function(df){
    m <- lm(y ~ x - 1, df);
    eq <- paste("Y = ", format(coef(m)[1], digits = 2), " * X , ", " r2 = ", format(summary(m)$r.squared, digits = 3), sep = "")
    return(eq)
}

```

Welcome
============================================

Summary {data-width=150}
--------------------------------------------

### About this Document

This document is an interactive dashboard viewable from most modern internet browsers. The dashboard is a validation and diagnostics tool for Activity Based Models. Users can compare model performance against a household survey as part of a validation exercise or compare two model runs for sensitivity testing. All of the data, charts, and maps viewable in this dashboard are embedded directly into the HTML file. An internet connection is necessary for the best user experience, but is not required.

Users may navigate to different areas of the dashboard using the navigation bar at the top of the page, and may interact directly with most tables, charts, and maps.

This document is best viewed using the most recent versions of the following web browsers:

* [Google Chrome](https://www.google.com/chrome/browser/desktop/)
* [Microsoft Internet Explorer](https://www.microsoft.com/en-us/download/internet-explorer.aspx)

Note: Mozilla Firefox does not correctly render the images in this HTML file.

Summary {data-width=600}
--------------------------------------------

### Modeling Region
```{r model_region}
# cat("Number of TAZs:", nrow(zone_shp))
# bins <- c(0, 10, 20, 50, 100, 200, 500, 1000, Inf)
# pal <- colorBin("YlOrRd", domain = zone_shp$HH, bins = bins)

m <- leaflet(data = zone_shp)%>%
 addTiles() %>%
 addProviderTiles(TILE_PROVIDER, group = "Background Map") %>%
 addLayersControl(
   overlayGroups = "Background Map", options = layersControlOptions(collapsed = FALSE)
 ) %>%
 addPolygons(weight = 0.5, opacity = 1)
m

# m <- leaflet(data = zone_shp)%>%
#  addTiles() %>%
#  addProviderTiles(TILE_PROVIDER, group = "Background Map") %>%
#  addLayersControl(
#    overlayGroups = "Background Map", options = layersControlOptions(collapsed = FALSE)
#  )
# m
#
```


Overview
============================================

Base Highlights {data-width=90}
--------------------------------------------

###

```{r Run_Date1_ValueBox}
sample_rate <- ifelse(IS_BASE_SURVEY=="Yes", "", as.character(BASE_SAMPLE_RATE*100))
valueBox(BASE_SCENARIO_NAME, paste("Sample Rate: ", sample_rate, "%", sep = ""), color = "DarkOrange")
base_pos <- which(base_csv_names=="totals")
base_df <- base_data[[base_pos]]
```

### Base Population
```{r Population1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_population"]/BASE_SAMPLE_RATE), big.mark = ","), "Population", icon = "ion-ios-people")
```

### Base Households
```{r Household1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_households"]/BASE_SAMPLE_RATE), big.mark = ","), "Households", icon = "glyphicon glyphicon-home")
```

### Base Tours
```{r Tours1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_tours"]/BASE_SAMPLE_RATE), big.mark = ","), "Total Tours", icon = "ion-refresh")
```

### Base Trips
```{r Trips1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_trips"]/BASE_SAMPLE_RATE), big.mark = ","), "Total Trips", icon = "ion-loop")
```

### Base Stops
```{r Stops1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_stops"]/BASE_SAMPLE_RATE), big.mark = ","), "Total Stops", icon = "ion-ios-location")
```

### Base VMT
```{r VMT1_ValueBox}
valueBox(prettyNum(round(base_df$value[base_df$name=="total_vmt"]/BASE_SAMPLE_RATE), big.mark = ","), "Total VMT", icon = "ion-android-car")
```



Build Highlights {data-width=90}
--------------------------------------------

###

```{r Run_Date2_ValueBox}
valueBox(BUILD_SCENARIO_NAME, paste("Sample Rate: ", BUILD_SAMPLE_RATE*100, "%", sep = ""), color = "DarkOrange")
build_pos <- which(build_csv_names=="totals")
build_df <- build_data[[build_pos]]
```

### Build Population
```{r Population2_ValueBox}
valueBox(prettyNum(round(build_df$value[build_df$name=="total_population"]/BUILD_SAMPLE_RATE), big.mark = ","), "Population", icon = "ion-ios-people")
```

### Build Households
```{r Household2_ValueBox}
valueBox(prettyNum(format(round(build_df$value[build_df$name=="total_households"]/BUILD_SAMPLE_RATE), scientific=F), big.mark = ","), "Households", icon = "glyphicon glyphicon-home")
```

### Build Tours
```{r Tours2_ValueBox}
valueBox(prettyNum(round(build_df$value[build_df$name=="total_tours"]/BUILD_SAMPLE_RATE), big.mark = ","), "Total Tours", icon = "ion-refresh")
```

### Build Trips
```{r Trips2_ValueBox}
valueBox(prettyNum(round(build_df$value[build_df$name=="total_trips"]/BUILD_SAMPLE_RATE), big.mark = ","), "Total Trips", icon = "ion-loop")
```

### Build Stops
```{r Stops2_ValueBox}
valueBox(prettyNum(round(build_df$value[build_df$name=="total_stops"]/BUILD_SAMPLE_RATE), big.mark = ","), "Total Stops", icon = "ion-ios-location")
```

### Build VMT
```{r VMT2_ValueBox}
valueBox(prettyNum(round(build_df$value[build_df$name=="total_vmt"]/BUILD_SAMPLE_RATE), big.mark = ","), "Total VMT", icon = "ion-android-car")
```


Chart Column 1 {data-width=275}
--------------------------------------------
### Person Type Distribution
```{r Chart_Person_Type}
base_pos <- which(base_csv_names=="pertypeDistbn")
base_df <- base_data[[base_pos]]
base_df$PERNAME <- person_type_df$name_char[match(base_df$PERTYPE, person_type_df$code)]
base_df$PERNAME <- factor(base_df$PERNAME, levels = person_type_char)
build_pos <- which(build_csv_names=="pertypeDistbn")
build_df <- build_data[[build_pos]]
build_df$PERNAME <- person_type_df$name_char[match(build_df$PERTYPE, person_type_df$code)]
build_df$PERNAME <- factor(build_df$PERNAME, levels = person_type_char)

colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "PERNAME", y = c("freq"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "Person Type", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, bottom_offset = 60, tickangle = -30)
p

```

### Household Size Distribution
```{r Chart_HHSize}
base_pos <- which(base_csv_names=="hhSizeDist")
base_df <- base_data[[base_pos]]

census_df = base_data[[which(base_csv_names=="hhSizeCensus")]]


build_pos <- which(build_csv_names=="hhSizeDist")
build_df <- build_data[[build_pos]]

colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "HHSIZE", y = c("freq"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "HH Size", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)
p

```

Base Highlights2 {data-width=90}
--------------------------------------------

###

```{r Run_Date3_ValueBox}
valueBox(BASE_SCENARIO_NAME, "", color = "DarkOrange")
base_pos <- which(base_csv_names=="totals")
base_df <- base_data[[base_pos]]
```


### Tours per Person
```{r TourRate3_Gauge}
rate <- base_df$value[base_df$name=="total_tours"]/base_df$value[base_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 2, gaugeSectors(danger = c(0,2), colors = c("Green", "Green", "Green")))
```

### Trips per Person
```{r TripRate3_Gauge}
rate <- base_df$value[base_df$name=="total_trips"]/base_df$value[base_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 5, gaugeSectors(danger = c(0,5), colors = c("Green", "Green", "Green")))
```

### Stops per Person
```{r StopRate3_Gauge}
rate <- base_df$value[base_df$name=="total_stops"]/base_df$value[base_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 2, gaugeSectors(danger = c(0,2), colors = c("Green", "Green", "Green")))
```

### Trips per Household
```{r TRate3_Gauge}
rate <- base_df$value[base_df$name=="total_trips"]/base_df$value[base_df$name=="total_households"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 15, gaugeSectors(danger = c(0,15), colors = c("Green", "Green", "Green")))
```


Build Highlights2 {data-width=90}
--------------------------------------------

###

```{r Run_Date4_ValueBox}
valueBox(BUILD_SCENARIO_NAME, "", color = "DarkOrange")
build_pos <- which(build_csv_names=="totals")
build_df <- build_data[[build_pos]]
```


### Tours per Person
```{r TourRate4_Gauge}
rate <- build_df$value[build_df$name=="total_tours"]/build_df$value[build_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 2, gaugeSectors(danger =  c(0,2), colors = c("Green", "Green", "Green")))
```

### Trips per Person
```{r TripRate4_Gauge}
rate <- build_df$value[build_df$name=="total_trips"]/build_df$value[build_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 5, gaugeSectors(danger = c(0,5), colors = c("Green", "Green", "Green")))
```

### Stops per Person
```{r StopRate4_Gauge}
rate <- build_df$value[build_df$name=="total_stops"]/build_df$value[build_df$name=="total_population"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 2, gaugeSectors(danger = c(0,2), colors = c("Green", "Green", "Green")))
```

### Trips per Household
```{r TRate4_Gauge}
rate <- build_df$value[build_df$name=="total_trips"]/build_df$value[build_df$name=="total_households"]
gauge(prettyNum(round(rate, 2), big.mark = ","), min = 0, max = 15, gaugeSectors(danger = c(0,15), colors = c("Green", "Green", "Green")))
```


Long Term Models{data-navmenu="Long Term"}
============================================

Description {.sidebar data-width=225}
--------------------------------------------


**Auto Ownership**

Results of household auto ownership model, which predicts number of vehicles per household.

**Mandatory TLFD**

Results of work and school location choice models.

Distribution of workers by distance between home and usual work place, and students by distance between home and school location.

Chart Column 1 {data-width=200}
--------------------------------------------
### Auto Ownership{data-height=265}
```{r Chart_Auto_Ownership}
if(IS_BASE_SURVEY=="Yes"){
  #cat("Census source: ", AO_CENSUS_LONG)
  # [3/31/2020] DH No census data used for SEMCOG
  base_df <- base_data[[which(base_csv_names=="autoOwnershipCensus")]]
}else{
  base_df <- base_data[[which(base_csv_names=="autoOwnership")]]
}

build_df <- build_data[[which(build_csv_names=="autoOwnershipCY")]]

build_df$X = NULL
base_df$X = NULL
#colnames(build_df) <- colnames(census_df)
#colnames(base_df) <- colnames(census_df)

sd.autoown <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "HHVEH", y = c("freq"), grp = "COUNTY", shared = TRUE)

p <- plotly_bar_plotter(data = sd.autoown, xlabel = "Number of Vehicles", ylabel = "Percent", ynames = c(AO_CENSUS_LONG, BUILD_SCENARIO_NAME), percent = T, height = 225)

bscols(widths=c(3,9),
  list(
    filter_select("ao_cname", "Select County", sd.autoown, ~grp_var, multiple=F)),
    p
  )


```

### Telecommute Frequency{data-height=250}
```{r Telecommute Frequency}

base_df <- base_data[[which(base_csv_names=="telecommuteFrequency")]]

build_df <- build_data[[which(build_csv_names=="telecommuteFrequency")]]

colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "telecommute_frequency", y = c("freq"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "Telecommute Frequency", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, height = 225, tickangle = -30, bottom_offset = 50)
p
```

### Percentage Working From Home{data-height=250}
```{r Chart_WFH}
base_df <- base_data[[which(base_csv_names=="wfh_summary")]]
base_df$share <- base_df$WFH/base_df$Workers
build_df <- build_data[[which(build_csv_names=="wfh_summary")]]
build_df$share <- build_df$WFH/build_df$Workers
std_DF <- cbind(base_df[,c("District", "share")], build_df[,c("share")])
colnames(std_DF) <- c("xvar", "prop1", "prop2")
p <- plotly_bar_plotter(data = std_DF, xlabel = "District", ylabel = "Percent WFH", ynames = c("Census ACS", BUILD_SCENARIO_NAME), percent = T, height = 275, tickangle = -320, bottom_offset = 25)
p

```



Chart Column 2 {data-width=350}
--------------------------------------------


### Mandatory TLFD{data-height=475}
```{r mandatoryTLFD}
base_df1 <- base_data[[which(base_csv_names=="workTLFD")]]
base_df1 <- melt(base_df1, id = c("distbin"))

base_df2 <- base_data[[which(base_csv_names=="univTLFD")]]
base_df2 <- melt(base_df2, id = c("distbin"))

base_df3 <- base_data[[which(base_csv_names=="schlTLFD")]]
base_df3 <- melt(base_df3, id = c("distbin"))

df_list = list(base_df1, base_df2, base_df3)
base_df = Reduce(function(x, y) merge(x, y, all=TRUE, by = c('distbin', 'variable')), df_list)
base_df[is.na(base_df)] = 0
# base_df <- cbind(base_df1, base_df2$value, base_df3$value)
colnames(base_df) <- c("distbin","variable","value1","value2","value3")

build_df1 <- build_data[[which(build_csv_names=="workTLFD")]]
build_df1 <- melt(build_df1, id = c("distbin"))

build_df2 <- build_data[[which(build_csv_names=="univTLFD")]]
build_df2 <- melt(build_df2, id = c("distbin"))

build_df3 <- build_data[[which(build_csv_names=="schlTLFD")]]
build_df3 <- melt(build_df3, id = c("distbin"))

build_df <- cbind(build_df1, build_df2$value, build_df3$value)
colnames(build_df) <- c("distbin","variable","value1","value2","value3")

sd.purpose <- get_sharedData(data_df1 = build_df, data_df2 = base_df, run1_name = BUILD_SCENARIO_NAME,
                             run2_name = BASE_SCENARIO_NAME, x = "distbin", y = c("value1", "value2", "value3"), grp = "variable")

p1 <- plotly_density_plotter(sd.purpose, index = "one", xlabel = "Miles to Work", percent = T, tickvals = seq(1,50,5), ticktext = seq(0,50,5), height = 240)
p2 <- plotly_density_plotter(sd.purpose, index = "two", xlabel = "Miles to University", percent = T, tickvals = seq(1,50,5), ticktext = seq(0,50,5), height = 240)
p3 <- plotly_density_plotter(sd.purpose, index = "three", xlabel = "Miles to School", percent = T, tickvals = seq(1,50,5), ticktext = seq(0,50,5), height = 240)

bscols(widths=c(12),
  list(filter_select("Purpose_County", "Select District", sd.purpose, ~grp_var,multiple=F),
  p1,
  p2,
  p3)
  )

```


Flows & Tour Lengths{data-navmenu="Long Term"}
============================================

Description {.sidebar data-width=135}
--------------------------------------------

**County-County Flow of Workers**

Crosstab of workers by home county and usual work place county.

Note: Districts can be Tract, County, District etc.

**Average Tour Lengths**

Average tour length to workplace by District of residence



Chart Column 1
--------------------------------------------

### {data-height=600}
```{r Table1_CountyFlows, eval  = TRUE}

# if(IS_BASE_SURVEY=="Yes"){
#   base_df <- base_data[[which(base_csv_names=="countyFlows")]]
#   # base_df <- base_data[[which(base_csv_names=="countyFlows")]]
#   name = 'Survey'
# }else{
#   base_df <- base_data[[which(base_csv_names=="countyFlowsCensus")]]
#   name = 'Census'
# }
base_df <- base_data[[which(base_csv_names=="countyFlows")]]
name = 'Survey'


base_df[,!colnames(base_df) %in% c("X")] <- base_df[,!colnames(base_df) %in% c("X")]/BASE_SAMPLE_RATE
if('X' %in% colnames(base_df)) {
  rownames(base_df) = base_df$X
  base_df$X = NULL
}
# t1 <- kable(base_df, format = 'html', caption = DISTRICT_FLOW_CENSUS, digits = 0, row.names = F, align = 'r', format.args = list(big.mark = ',')) %>%
#   kable_styling('striped', font_size = 10)
# t1

base_df2 = base_df[1:(nrow(base_df) - 1), 1:(ncol(base_df) - 1)]

nworkers = formatC(sum(base_df2), format="f", big.mark=",", digits=0)
cat("County - County Flow of Workers -", name, ':', nworkers, 'workers in region')


diag = chorddiag(as.matrix(base_df2), palette = 'Spectral')

diag

```

### {data-height=350}
```{r Table1_MandTripLengths}
cat("Average Mandatory Tour Lengths")

base_df <- base_data[[which(base_csv_names=="mandTripLengths")]]
df <- base_df
colnames(df) <- c("Home District", "Work","University","School")

eval_expr <- paste("t1 <- kable(df, format = 'html', digits = 2, row.names = F, align = 'r', format.args = list(big.mark = ',')) %>%
  kable_styling('striped', font_size = 10, full_width=F, position='center') %>%
  add_header_above(c(' ' = 1, '", BASE_SCENARIO_NAME, "' = 3))", sep = "")
eval(parse(text = eval_expr))
t1
```


Chart Column 2
--------------------------------------------
### {data-height=600}
```{r Table2_CountyFlows, eval = TRUE}


build_pos <- which(build_csv_names=="countyFlows")
build_df <- build_data[[build_pos]]

#build_df <- base_data[[which(base_csv_names=="countyFlowsCensus")]]
build_df[,!colnames(build_df) %in% c("X")] <- build_df[,!colnames(build_df) %in% c("X")]/BUILD_SAMPLE_RATE
# t2 <- kable(build_df, format = 'html', caption = BUILD_SCENARIO_NAME, digits = 0, row.names = F, align = 'r', format.args = list(big.mark = ',')) %>%
#   kable_styling('striped', font_size = 10)
# t2
if('X' %in% colnames(build_df)) {
  rownames(build_df) = build_df$X
  build_df$X = NULL
}

build_df2 = build_df[1:(nrow(build_df) - 1), 1:(ncol(build_df) - 1)]

nworkers = formatC(sum(build_df2), format="f", big.mark=",", digits=0)
#cat("County - County Flow of Workers - ACS:", nworkers, 'workers in region')
cat("County - County Flow of Workers - Model:", nworkers, 'workers in region')


diag2 = chorddiag(as.matrix(build_df2), palette = 'Spectral')

diag2

```

### {data-height=350}
```{r Table2_MandTripLengths}
cat("Average Mandatory Tour Lengths")

build_df <- build_data[[which(build_csv_names=="mandTourLengths")]]
df <- build_df
colnames(df) <- c("Home District", "Work","University","School")

eval_expr <- paste("t2 <- kable(df, format = 'html', digits = 2, row.names = F, align = 'r', format.args = list(big.mark = ',')) %>%
  kable_styling('striped', font_size = 10, full_width=F, position='center') %>%
  add_header_above(c(' ' = 1, '", BUILD_SCENARIO_NAME, "' = 3))", sep = "")
eval(parse(text = eval_expr))
t2
```

Employment vs Workers{data-navmenu="Long Term"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Employment vs Workers comparison at TAZ level**

Results of work location model.

Comparison of assigned workers to available employment at TAZ level.

Only for build scenario.


Chart Column 2{.tabset}
--------------------------------------------

### By County{data-height=575}

```{r job_wrk1a}
base_df <- build_data[[which(build_csv_names=="employees_per_co")]]
build_df <- build_data[[which(build_csv_names=="workers_per_co")]]

# base_df$WDISTRICT = levels(base_df$WDISTRICT)[base_df$WDISTRICT]
# build_df$WDISTRICT = levels(build_df$WDISTRICT)[build_df$WDISTRICT]
build_df = build_df[build_df$WDISTRICT %in% base_df$WDISTRICT,]
base_df = base_df[base_df$WDISTRICT %in% build_df$WDISTRICT,]

#colnames(base_df)[which(colnames(base_df) == "freq")] = BASE_SCENARIO_NAME
#colnames(build_df)[which(colnames(build_df) == "freq")] = BUILD_SCENARIO_NAME
# 
df = base_df
df$freq_build = build_df$freq[which(df$WDISTRICT == build_df$WDISTRICT)]/BUILD_SAMPLE_RATE

co_cor = cor(df$freq, df$freq_build)^2
co_lm = lm(freq_build ~ freq, df)

text1 = paste("r2 =", round(co_cor, 4))
text2 = paste("y = ", round(co_lm$coefficients[['freq']], 4), "x  + ", round(co_lm$coefficients[['(Intercept)']], 2), sep = "")

ggplot(df, aes(x = freq, y = freq_build, label = WDISTRICT)) + geom_point() +
  geom_abline(slope = 1, intercept = 0, color = "red") +
  geom_text(vjust = 0, nudge_y = 0.1, check_overlap = TRUE) +
  scale_x_log10("SEDATA Input", labels = comma, limits = c(10000, max(df$freq, df$build_freq))) +
  scale_y_log10("ActivitySim", labels = comma, limits = c(10000, max(df$freq, df$build_freq))) +
  annotate("text", x=500000, y=c(20000, 15000), label= c(text1, text2))
```

### By TAZ{data-height=575}
```{r job_wrk_taz}
base_df <- build_data[[which(build_csv_names=="employees_per_taz")]]
build_df <- build_data[[which(build_csv_names=="workers_per_taz")]]


df = data.frame(taz = 1:3632) %>%
  dplyr::left_join(base_df, by = "taz") %>%
  dplyr::left_join(build_df, by = c("taz" = "worktaz")) %>%
  rename(base = freq.x) %>%
  mutate(build = freq.y/BUILD_SAMPLE_RATE) %>%
  mutate(build = ifelse(is.na(build), 0, build),
         base = ifelse(is.na(base), 0, base))

co_cor = cor(df$base, df$build)^2
co_lm = lm(build ~ base, df)

text1 = paste("r2 =", round(co_cor, 4))
text2 = paste("y = ", round(co_lm$coefficients[['base']], 4), "x  + ", round(co_lm$coefficients[['(Intercept)']], 2), sep = "")

ggplot(df, aes(x = base, y = build)) + geom_point(alpha = 0.3) +
  geom_abline(slope = 1, intercept = 0, color = "red") +
  scale_x_continuous("SEDATA Input", labels = comma, limits = c(0, max(df$freq, df$build_freq))) +
  scale_y_continuous("ActivitySim", labels = comma, limits = c(0, max(df$freq, df$build_freq))) +
  annotate("text", x=60000, y=c(4000, 1500), label= c(text1, text2))
```

### Low Income{data-height=575}
```{r job_wrk2}

#knitr::include_graphics(file.path(SYSTEM_JPEG_PATH, "workersByEmpCat_inc1.jpeg"))


# inc1_plots <- paste0(list.files(SYSTEM_JPEG_PATH, "inc1_", full.names=TRUE))
# # inc1_plot_names <- str_replace_all(inc1_plots, c("workers_inc1_" = "", SYSTEM_JPEG_PATH = "", ".jpeg"= ""))
# cat(inc1_plots)
# cat(SYSTEM_JPEG_PATH)
# selectInput('workers_inc1', 'Employment Category', inc1_plots, selected="total")
# renderImage(input$workers_inc1, deleteFile=FALSE)


# bsselect(inc1_plots, type = "img", selected = "tot", live_search = TRUE, show_tick = TRUE)
```

### Medium Income{data-height=575}
```{r job_wrk3}

#knitr::include_graphics(file.path(SYSTEM_JPEG_PATH, "workersByEmpCat_inc2.jpeg"))

```

### High Income{data-height=575}
```{r job_wrk4}

#knitr::include_graphics(file.path(SYSTEM_JPEG_PATH, "workersByEmpCat_inc3.jpeg"))

```

### Very High Income{data-height=575}
```{r job_wrk5}

#knitr::include_graphics(file.path(SYSTEM_JPEG_PATH, "workersByEmpCat_inc4.jpeg"))

```

### Total{data-height=575}
```{r job_wrk7}

#knitr::include_graphics(file.path(SYSTEM_JPEG_PATH, "workers_incAll.jpeg"))

```


Zero Auto Households {data-navmenu="Long Term"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Census vs Model comparison at Census Tract level**

Compares number of zero auto households.

Census Data is from 2017 five year ACS.

Only for build scenario.

Summary
--------------------------------------------

### Zero Auto Households Census vs Model
```{r zero_auto_hh}

# BEWARE! shp data names are shortend by writeOGR in the script that creates the ct_zero_auto_shp file
#  Names in that script won't match names shown here.
#  For example, ct_zero_auto_shp@data$Diff_ZeroAuto ->  ct_zero_auto_shp@data$Dff_ZrA
labels <- sprintf(
  "%s
%s %.2f %s
 %s
Census 0A: %s%s, Model 0A: %s%s", ct_zero_auto_shp@data$TRACTCE, ct_zero_auto_shp@data$txtCmm2, ct_zero_auto_shp@data$Dff_ZrA, "%", ct_zero_auto_shp@data$txtCmm1, round(ct_zero_auto_shp@data$Cns_A0P, 2), "%", round(ct_zero_auto_shp@data$Mdl_A0P, 2), "%") %>% lapply(htmltools::HTML)

bins <- c(-Inf, -100, -75, -50, -25, -5, 5, 25, 50, 75, 100, Inf)
pal <- colorBin("PiYG", domain = ct_zero_auto_shp@data$Dff_ZrA, na.color="transparent", bins = bins)

m <- leaflet(data = ct_zero_auto_shp)%>%
  addTiles() %>%
  addProviderTiles(TILE_PROVIDER, group = "Background Map") %>%
  addLayersControl(
    overlayGroups = "Background Map", options = layersControlOptions(collapsed = FALSE)
  ) %>%
  addPolygons(group='ZeroCarDiff',
              fillColor = ~pal(Dff_ZrA),
              weight = 0.2,
              opacity = 1,
              color = "gray",
              stroke=T,
              dashArray = "5, 1",
              fillOpacity = 0.7,
              highlight = highlightOptions(
                weight = 1,
                color = "blue",
                dashArray = "",
                fillOpacity = 0.7,
                bringToFront = TRUE),
              label = labels,
              labelOptions = labelOptions(
                style = list("font-weight" = "normal", padding = "3px 8px"),
                textsize = "15px",
                direction = "auto")) %>%
  addLegend(pal = pal, values = ~density, opacity = 0.7, title = "Estimated(%) - Observed(%) Bins",
            position = "bottomright")
m
```

Tour Summaries{data-navmenu="Tour Level"}
============================================

Description {.sidebar data-width=225}
--------------------------------------------

This page summarizes day-pattern and tour generation model results.

**Daily Activity Pattern**

Results of Coordinated Daily Activity Pattern (CDAP) model, summarized for each person.

_M_ : One or more mandatory tours

_N_ : No mandatory tours but one or more non-mandatory tours

_H_ : No tours (either home all day or out of area)

**Percentage of Households with Joint Tour**

Also the result of the CDAP model, summarized for each household.

**Mandatory Tour Frequency**

Result of the mandatory tour frequency model, summarized for each person with a daily activity pattern type _M_

**Tour rate by person type**

Summary of tours per person resulting from all tour generation models. Joint tours are counted for each participant.

**Individual non-mandatory tour frequency**

Results of individual non-mandatory tour frequency model, summarized for each person with a daily activity pattern type _M_ or _N_.

Chart Column 1 {data-width=160}
--------------------------------------------

### Daily Activity Pattern{data-height=500}
```{r Hist_DAP}
base_df <- base_data[[which(base_csv_names=="dapSummary_vis")]]
base_df$PERNAME <- person_type_df$name_char[match(base_df$PERTYPE, person_type_df$code)]
base_df$PERNAME <- factor(base_df$PERNAME, levels = person_type_char)
base_df$DAP <- factor(base_df$DAP, levels = dap_types)
build_df <- build_data[[which(build_csv_names=="dapSummary_vis")]]
build_df$PERNAME <- person_type_df$name_char[match(build_df$PERTYPE, person_type_df$code)]
build_df$PERNAME <- factor(build_df$PERNAME, levels = person_type_char)
build_df$DAP <- factor(build_df$DAP, levels = dap_types)

base_df$grp <- BASE_SCENARIO_NAME
build_df$grp <- BUILD_SCENARIO_NAME
colnames(build_df) <- colnames(base_df)

sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="DAP", y = c("freq"), grp = "PERNAME", shared = T)
p <- plotly_bar_plotter(data = sd.pertype, height = 250, xlabel = "DAP", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)

bscols(widths=c(3,9),
  list(
    filter_select("pertype_dap", "Select Person Type", sd.pertype, ~grp_var,multiple=F)),
    p
  )

```

### Percentage of Households with a Joint Tour{data-height=300}
```{r Hist_Presence_Joint}
base_pos <- which(base_csv_names=="hhsizeJoint")
base_df <- base_data[[base_pos]]
base_df <- base_df %>%
  group_by(HHSIZE) %>%
  mutate(percent = prop.table(freq)) %>%
  filter(JOINT==1) %>%
  ungroup()
build_pos <- which(build_csv_names=="hhsizeJoint")
build_df <- build_data[[build_pos]]
build_df <- build_df %>%
  group_by(HHSIZE) %>%
  mutate(percent = prop.table(freq)) %>%
  filter(JOINT==1) %>%
  ungroup()

colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "HHSIZE", y = c("percent"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "HH Size", ylabel = "Percent of Households", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = F, tickformat = ".1%", hoverformat = ".1%")
p

```

### Mandatory Tour Frequency{data-height=500}
```{r Hist_MTF}
base_pos <- which(base_csv_names=="mtfSummary_vis")
base_df <- base_data[[base_pos]]
base_df$PERNAME <- person_type_df$name_char[match(base_df$PERTYPE, person_type_df$code)]
base_df$PERNAME <- factor(base_df$PERNAME, levels = person_type_char)
base_df$mtf_name <- mtf_df$name[match(base_df$MTF, mtf_df$code)]
base_df$mtf_name <- factor(base_df$mtf_name, levels = mtf_names)
build_pos <- which(build_csv_names=="mtfSummary_vis")
build_df <- build_data[[build_pos]]
build_df$PERNAME <- person_type_df$name_char[match(build_df$PERTYPE, person_type_df$code)]
build_df$PERNAME <- factor(build_df$PERNAME, levels = person_type_char)
build_df$mtf_name <- mtf_df$name[match(build_df$MTF, mtf_df$code)]
build_df$mtf_name <- factor(build_df$mtf_name, levels = mtf_names)
colnames(build_df) <- colnames(base_df)

sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="mtf_name", y = c("freq"), grp = "PERNAME", shared = T)
p <- plotly_bar_plotter(data = sd.pertype, height = 250, xlabel = "MTF Choice", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickangle = -30, bottom_offset = 50)

bscols(widths=c(3,9),
  list(
    filter_select("pertype_mtf", "Select Person Type", sd.pertype, ~grp_var,multiple=F)),
    p
  )

```

Chart Column 1 {data-width=150}
--------------------------------------------
### Total Tour Rate (only active Persons)
```{r Hist_totaltours}
base_df <- base_data[[which(base_csv_names=="total_tours_by_pertype_vis")]]
base_df$PERNAME <- person_type_df$name_char[match(base_df$PERTYPE, person_type_df$code)]
base_df$PERNAME <- factor(base_df$PERNAME, levels = person_type_char)
base_df1 <- base_data[[which(base_csv_names=="activePertypeDistbn")]]
base_df$persons <- base_df1$freq[match(base_df$PERTYPE, base_df1$PERTYPE)]
base_df$tourrate <- round(base_df$freq/base_df$persons,2)

build_df <- build_data[[which(build_csv_names=="total_tours_by_pertype_vis")]]
build_df$PERNAME <- person_type_df$name_char[match(build_df$PERTYPE, person_type_df$code)]
build_df$PERNAME <- factor(build_df$PERNAME, levels = person_type_char)
build_df1 <- build_data[[which(build_csv_names=="activePertypeDistbn")]]
build_df$persons <- build_df1$freq[match(build_df$PERTYPE, build_df1$PERTYPE)]
build_df$tourrate <- round(build_df$freq/build_df$persons,2)

colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "PERNAME", y = c("tourrate"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "Person Type", ylabel = "Tour Rate", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = F, height = 340, tickangle = -30, bottom_offset = 50)
p


```


### Persons by Individual Non-Mandatory Tours
```{r Hist_INM}
base_df <- base_data[[which(base_csv_names=="inmSummary_vis")]]
base_df$PERNAME <- person_type_df$name_char[match(base_df$PERTYPE, person_type_df$code)]
base_df$PERNAME <- factor(base_df$PERNAME, levels = person_type_char)

build_df <- build_data[[which(build_csv_names=="inmSummary_vis")]]
build_df$PERNAME <- person_type_df$name_char[match(build_df$PERTYPE, person_type_df$code)]
build_df$PERNAME <- factor(build_df$PERNAME, levels = person_type_char)

colnames(build_df) <- colnames(base_df)

sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="nmtours", y = c("freq"), grp = "PERNAME", shared = T)
#p <- plotly_bar_plotter(data = sd.pertype, height = 340, xlabel = "Number of Tours", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, #BUILD_SCENARIO_NAME), percent = T, tickvals = c(seq(0,2), "3pl"), ticktext = c("0", "1", "2", "3pl"))

p <- plotly_bar_plotter(data = sd.pertype, height = 340, xlabel = "Number of Tours", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)

bscols(widths=c(3,9),
  list(
    filter_select("pertype_mtf", "Select Person Type", sd.pertype, ~grp_var,multiple=F)),
    p
  )


```



Joint Tours{data-navmenu="Tour Level"}
============================================

Description {.sidebar data-width=225}
--------------------------------------------

********

This page tabulates the results of the Joint Tour Frequency and Composition Model and the Joint Tour Person Participation Model.

**Joint Tour Frequency**

The frequency of households by number and purpose of joint tours.

**Joint Tour Composition**

The frequency of tours by composition (Adults only, Children only, Adults + Children).

**Joint Tour Party Size**

The frequency of joint tours by the number of household members participating in the tour.

**Joint Tours by HH Size**

The frequency of households by household size and the number of joint tours per household.

**Joint Tours by HH Size**

_Tour Level_

Distribution of joint tours by party size for each composition type.


Chart Column 1 {data-width=150}
--------------------------------------------
### Joint Tour Frequency{data-height=675}
```{r jtf}
base_df <- base_data[[which(base_csv_names=="jtf")]]
build_df <- build_data[[which(build_csv_names=="jtf")]]
# remove no joint tours option
base_df <- base_df[-1,]
build_df <- build_df[-1,colnames(base_df)]
colnames(build_df) <- colnames(base_df)

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "alt_name", y = c("freq"))
std_DF$xvar <- factor(std_DF$xvar, levels = jtf_alternatives)

p <- plotly_bar_plotter(data = std_DF, xlabel = "Joint Tour Combination", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, height = 500, bottom_offset = 275, tickangle = 300)
p

```

### Joint Tour Composition
```{r jtf_comp}
base_df <- base_data[[which(base_csv_names=="jointComp")]]
names(base_df)[names(base_df)=="tour_composition"] <- "COMPOSITION"
build_df <- build_data[[which(build_csv_names=="jointComp")]]
colnames(build_df) <- colnames(base_df)

p1 <- plotly_pie_chart(data = base_df, label_var = "COMPOSITION", value_var = "freq", height = 250, title = BASE_SCENARIO_NAME, top_offset = 50)
p2 <- plotly_pie_chart(data = build_df, label_var = "COMPOSITION", value_var = "freq", height = 250, title = BUILD_SCENARIO_NAME, top_offset = 50)

bscols(widths=c(6,6),
  p1,
  p2
  )

```

Chart Column 1 {data-width=150}
--------------------------------------------

### Joint Tours By Number of Household Members
```{r jtf_partysize}
base_df <- base_data[[which(base_csv_names=="jointPartySize")]]
build_df <- build_data[[which(build_csv_names=="jointPartySize")]]
colnames(build_df) <- colnames(base_df)

build_df$freq[build_df$NUMBER_HH==5] <- sum(build_df$freq[build_df$NUMBER_HH>=5])
build_df <- build_df[build_df$NUMBER_HH<=5, ]

std_DF <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "NUMBER_HH", y = c("freq"))
p <- plotly_bar_plotter(data = std_DF, xlabel = "Joint Party Size", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, height = 200)
p

```

### Joint Tours by Household Size
```{r jtf_byhhsize}
base_pos <- which(base_csv_names=="jointToursHHSize")
base_df <- base_data[[base_pos]]

build_pos <- which(build_csv_names=="jointToursHHSize")
build_df <- build_data[[build_pos]]
colnames(build_df) <- colnames(base_df)

sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="jointTours", y = c("freq"), grp = "hhsize", shared = T)
p <- plotly_bar_plotter(data = sd.pertype, height = 225, xlabel = "Number of Joint Tours", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)

bscols(widths=c(3,9),
  list(
    filter_select("jtf_hhsize", "Select HH Size Group", sd.pertype, ~grp_var,multiple=F)),
    p
  )

```

### Party Size Distribution by Joint Tour Composition
```{r jtf_comppartysize}
base_df <- base_data[[which(base_csv_names=="jointCompPartySize")]]
build_df <- build_data[[which(build_csv_names=="jointCompPartySize")]]
colnames(build_df) <- colnames(base_df)

sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="partysize", y = c("freq"), grp = "comp", shared = T)
p <- plotly_bar_plotter(data = sd.pertype, height = 225, xlabel = "Joint Party Size", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)

bscols(widths=c(3,9),
  list(
    filter_select("jtf_comp", "Select Party Composition", sd.pertype, ~grp_var,multiple=F)),
    p
  )

```




Destination{data-navmenu="Tour Level"}
============================================

Description {.sidebar data-width=225}
--------------------------------------------

********

**Non-Mandatory Tour Length Distribution**

Results of non-mandatory tour destination choice models.

Distribution of tours by distance between tour origin and destination for each non-mandatory tour purpose.


Chart Column 1 {data-width=100}
--------------------------------------------
### Non-Mandatory Tour Length Distribution{data-height=350}
```{r nm_tlfd}
base_df <- base_data[[which(base_csv_names=="tourDistProfile_vis")]]
build_df <- build_data[[which(build_csv_names=="tourDistProfile_vis")]]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$PURPOSE <- as.character(base_df$PURPOSE)
build_df$PURPOSE <- as.character(build_df$PURPOSE)
base_df$PURPOSE <- purpose_type_df$name[match(base_df$PURPOSE, purpose_type_df$code)]
build_df$PURPOSE <- purpose_type_df$name[match(build_df$PURPOSE, purpose_type_df$code)]

sd.purpose <- get_sharedData(data_df1 = base_df, data_df2 = build_df, run1_name = BASE_SCENARIO_NAME,
                             run2_name = BUILD_SCENARIO_NAME, x = "distbin", y = c("freq"), grp = "PURPOSE")

p1 <- plotly_density_plotter(sd.purpose, index = "one", xlabel = "Miles", percent = T,
                             tickvals = seq(2,41), ticktext = c(seq(1,40), "40pl"))
bscols(widths=c(2,10),
  filter_select("Tour Purpose", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F),
  p1
  )
```

### Average Non-Mandatory Tour Lengths (Miles){data-height=250}
```{r Table1_nonMandTripLength}
base_df <- base_data[[which(base_csv_names=="nonMandTripLengths")]]
build_df <- build_data[[which(build_csv_names=="nonMandTourLengths")]]
df <- data.frame(base_df, build_df[,-1])
colnames(df) <- c("Purpose", BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME)
df$Purpose <- purpose_type_df$name[match(df$Purpose, purpose_type_df$code)]

t1 <- kable(df, format = "html", digits = 2, row.names = F, align = 'c', format.args = list(big.mark = ',')) %>%
  kable_styling("striped", full_width = F)
t1
```


TOD {data-navmenu="Tour Level"}
============================================

Description {.sidebar data-width=200}
--------------------------------------------

********

**Tour Departure Arrival & Duration**

Tour Time-of-day Choice Model results.

Each tour is assigned a time period of departure (time leaving home or work) and arrival (time arriving back at home or work). The entire day is divided into 18 one-hour bins (the first bin includes 3:00 AM to 6:00 AM and the last bin includes 11:00 PM to 3:00 AM).

Tour duration is calculated as a function of departure and arrival period. It includes travel time and time spent at the primary destination and all intermediate stops.

Results are shown for tours, filtered by tour purpose.

********













Chart Column 1 {.tabset}
--------------------------------------------
### Tour Departure-Arrival Profile
```{r tour_tod}
base_df <- base_data[[which(base_csv_names=="todProfile_vis")]]
base_df$tod_bin <- tod_df$bin[match(base_df$id, tod_df$id)]
base_df$dur_bin <- dur_df$bin[match(base_df$id, dur_df$id)]
build_df <- build_data[[which(build_csv_names=="todProfile_vis")]]
build_df$tod_bin <- tod_df$bin[match(build_df$id, tod_df$id)]
build_df$dur_bin <- dur_df$bin[match(build_df$id, dur_df$id)]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.purpose <- get_sharedData(data_df1 = base_df, data_df2 = build_df, run1_name = BASE_SCENARIO_NAME,
                             run2_name = BUILD_SCENARIO_NAME, x = "id", y = c("freq_dep", "freq_arr", "freq_dur"), grp = "purpose")

p1 <- plotly_density_plotter(sd.purpose, index = "one", xlabel = "Tour Departure", percent = T, left_offset = 25,
                             tickvals = seq(1,48), ticktext = todBins, bottom_offset = 150, tickangle = 315, height = 275)
p2 <- plotly_density_plotter(sd.purpose, index = "two", xlabel = "Tour Arrival", percent = T, left_offset = 25,
                             tickvals = seq(1,48), ticktext = todBins, bottom_offset = 150, tickangle = 315, height = 275)
p3 <- plotly_density_plotter(sd.purpose, index = "three", xlabel = "Tour Duration", percent = T, left_offset = 25,
                             tickvals = seq(1,48), ticktext = durBins, bottom_offset = 50, tickangle = 315, height = 225)

bscols(widths=c(2,10),
  filter_select("Tour Purpose", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F),
  list(p1, p2, p3)
  )


```

### Tour Aggregate Departure-Arrival Profile
```{r tour_tod_agg}
base_df <- base_data[[which(base_csv_names=="todProfile_vis")]]
base_df$tod_agg <- cut(base_df$id, breaks = timePeriodBreaks, labels = timePeriods, right = FALSE)
base_df$tod_agg <- factor(base_df$tod_agg, levels = timePeriodOrder)
base_df <- base_df %>%
  dplyr::group_by(purpose, tod_agg) %>%
  dplyr::summarise(freq_dep = sum(freq_dep), freq_arr = sum(freq_arr), freq_dur = sum(freq_dur)) %>%
  dplyr::ungroup()
base_df = base_df %>% arrange(match(tod_agg, timePeriodOrder))

build_df <- build_data[[which(build_csv_names=="todProfile_vis")]]
build_df$tod_agg <- cut(build_df$id, breaks = timePeriodBreaks, labels = timePeriods, right = FALSE)
build_df$tod_agg <- factor(build_df$tod_agg, levels = timePeriodOrder)

build_df <- build_df %>%
  dplyr::group_by(purpose, tod_agg) %>%
  dplyr::summarise(freq_dep = sum(freq_dep), freq_arr = sum(freq_arr), freq_dur = sum(freq_dur)) %>%
  dplyr::ungroup()
build_df = build_df %>% arrange(match(tod_agg, timePeriodOrder))

colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]

sd.purpose <- get_standardDF(data_df1=base_df, data_df2=build_df, x="tod_agg", y = c("freq_dep", "freq_arr", "freq_dur"), grp = "purpose", shared = T)

p1 <- plotly_bar_plotter(data = sd.purpose, height = 350, xlabel = "Tour Departure", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)
p2 <- plotly_bar_plotter(data = sd.purpose, height = 350, xlabel = "Tour Arrival", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, index = 2)

bscols(widths=c(2,10),
  filter_select("Tour Purpose", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F),
  list(p1, p2)
  )


```



Tour Mode{data-navmenu="Tour Level"}
============================================


Chart Column 1{data-width=150}
--------------------------------------------


### Tour Mode Choice
```{r tourMode}
 # [3/31/2020] DH
# base_df <- base_data[[which(base_csv_names=="tmodeProfile_vis_CHTS_OBS")]]
if(IS_BASE_SURVEY=="Yes"){
  base_df <- base_data[[which(base_csv_names=="tmodeProfile_vis_HTS")]]
}else{
  base_df <- base_data[[which(base_csv_names=="tmodeProfile_vis")]]
}
base_df$purpose <- as.character(base_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
if(BASE_SCENARIO_NAME == "SEMCOG_HTS") {
  build_df <- build_data[[which(build_csv_names=="tmodeProfile_vis")]]
} else {
  build_df <- build_data[[which(build_csv_names=="tmodeProfile_vis")]]
}


# build_df <- build_data[[which(build_csv_names=="tmodeProfile_vis_CHTS")]]
build_df$purpose <- as.character(build_df$purpose)
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]
colnames(build_df) <- colnames(base_df)


sd.pertype <- get_standardDF(data_df1=base_df, data_df2=build_df, x="id", y = c("freq_as0", "freq_as1", "freq_as2", "freq_all"), grp = "purpose", shared = T)
p1 <- plotly_bar_plotter(data = sd.pertype, height = 375, xlabel = "Tour Mode [Zero Auto]", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,length(tourMode)), ticktext = tourMode, bottom_offset = 55, tickangle = 300)
p2 <- plotly_bar_plotter(data = sd.pertype, height = 375, xlabel = "Tour Mode [Autos < Workers]", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,length(tourMode)), ticktext = tourMode, index = 2, bottom_offset = 55, tickangle = 300)
p3 <- plotly_bar_plotter(data = sd.pertype, height = 375, xlabel = "Tour Mode [Autos >= Workers]", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,length(tourMode)), ticktext = tourMode, index = 3, bottom_offset = 55, tickangle = 300)
p4 <- plotly_bar_plotter(data = sd.pertype, height = 375, xlabel = "Tour Mode [Total]", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,length(tourMode)), ticktext = tourMode, index = 4, bottom_offset = 55, tickangle = 300)

filter_select("tourMode", "Select Tour Purpose", sd.pertype, ~grp_var,multiple=F)

```

********


**Tour Mode Choice**

Results of Tour Mode Choice Models, which selects a primary mode for each tour.

Distribution of tours by tour mode and the ratio of autos to drivers in the household.


Chart Column 2 {data-width=400}
--------------------------------------------

###
```{r tourMode2}
bscols(widths=c(12),
  list(p1,p2)
  )
```

Chart Column 3 {data-width=400}
--------------------------------------------

###
```{r tourMode3}
bscols(widths=c(12),
  list(p3,p4)
  )
```


Stop Frequency {data-navmenu="Trip Level"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Stop Frequency**

Results of the Intermediate Stop Frequency Model, which predicts the number of intermediate stops on each tour by tour direction (outbound versus inbound).

The summary shows percent of tours by number of stops on the tour and tour direction.

**Stop Purpose**

Results of the Intermediate Stop Purpose Model, which is currently implemented as a Monte Carlo choice according to probability distributions generated from survey data.

The summary shows the percent of intermediate stops by stop purpose and tour purpose.

Chart Column 1 {data-width=200}
--------------------------------------------
### Stop Frequency - Directional
```{r stopfreq_dir}
base_df <- base_data[[which(base_csv_names=="stopfreqDir_vis")]]
build_df <- build_data[[which(build_csv_names=="stopfreqDir_vis")]]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.pertype1 <- get_standardDF(data_df1=base_df, data_df2=build_df, x="nstops", y = c("freq_out", "freq_inb"), grp = "purpose", shared = T)
p1 <- plotly_bar_plotter(data = sd.pertype1, height = 325, xlabel = "Number of Stops - Outbound", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,4), ticktext = c("0", "1", "2", "3pl"))
p2 <- plotly_bar_plotter(data = sd.pertype1, height = 325, xlabel = "Number of Stops - Inbound", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,4), ticktext = c("0", "1", "2", "3pl"), index = 2)
bscols(widths=c(12),
  list(
    filter_select("stopfreq_dir", "Select Tour Purpose", sd.pertype1, ~grp_var,multiple=F),
    p1,
    p2)
  )

```



Chart Column 1 {data-width=300}
--------------------------------------------
### Stop Frequency - Total{data-height=250}
```{r stopfreq_total}
base_df <- base_data[[which(base_csv_names=="stopfreq_total_vis")]]
build_df <- build_data[[which(build_csv_names=="stopfreq_total_vis")]]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.pertype2 <- get_standardDF(data_df1=base_df, data_df2=build_df, x="nstops", y = c("freq"), grp = "purpose", shared = T)
p1 <- plotly_bar_plotter(data = sd.pertype2, height = 350, xlabel = "Number of Stops", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,7), ticktext = c("0", "1", "2", "3", "4", "5", "6pl"))

bscols(widths=c(3,9),
  list(
    filter_select("stopfreq_total", "Select Tour Purpose", sd.pertype2, ~grp_var,multiple=F)),
    p1
  )
```

### Stop Purpose by Tour Purpose{data-height=250}
```{r stoppurp_tourpurp}
base_df <- base_data[[which(base_csv_names=="stoppurpose_tourpurpose_vis")]]
build_df <- build_data[[which(build_csv_names=="stoppurpose_tourpurpose_vis")]]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.pertype3 <- get_standardDF(data_df1=base_df, data_df2=build_df, x="stop_purp", y = c("freq"), grp = "purpose", shared = T)
p1 <- plotly_bar_plotter(data = sd.pertype3, height = 350, xlabel = "Stop Purpose", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,9), ticktext = stopPurposes[1:9])

bscols(widths=c(3,9),
  list(
    filter_select("stoppurp_tourpurp", "Select Tour Purpose", sd.pertype3, ~grp_var,multiple=F)),
    p1
  )

```


Location{data-navmenu="Trip Level"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Stop Location**

Results of the Intermediate Stop Location Choice Model, which predicts the location of each intermediate stop.

The summary shows the distribution of intermediate stops by out of direction distance and tour purpose.

Out of direction distance is defined as the extra distance to the destination as a result of traveling through the stop location.
For stops in the outbound direction, it is based on the distance between the last known location (the tour origin or previous outbound stop) and the tour primary destination.
For stops in the inbound direction, it is based on the distance between the last known location (the tour primary destination or previous inbound stop) and the tour origin.

Chart Column 1 {data-width=800}
--------------------------------------------

### Stop Location - Out of Direction Distance{data-height=350}
```{r stopDC}
base_df <- base_data[[which(base_csv_names=="stopDC_vis")]]
build_df <- build_data[[which(build_csv_names=="stopDC_vis")]]
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$PURPOSE <- as.character(base_df$PURPOSE)
build_df$PURPOSE <- as.character(build_df$PURPOSE)
base_df$PURPOSE <- purpose_type_df$name[match(base_df$PURPOSE, purpose_type_df$code)]
build_df$PURPOSE <- purpose_type_df$name[match(build_df$PURPOSE, purpose_type_df$code)]


sd.purpose <- get_sharedData(data_df1 = base_df, data_df2 = build_df, run1_name = BASE_SCENARIO_NAME,
                             run2_name = BUILD_SCENARIO_NAME, x = "distbin", y = c("freq"), grp = "PURPOSE")

p1 <- plotly_density_plotter(sd.purpose, index = "one", xlabel = "Out of Direction Distance (Miles)", percent = T, left_offset = 25,
                             tickvals = seq(1,42), ticktext = outDirDist, height = 600, tickangle = 300, bottom_offset = 50)
bscols(widths=c(12),
  list(
    filter_select("stopDC", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F), p1)
  )
```

Chart Column 1 {data-width=300}
--------------------------------------------

### Average Out of Direction Distance (Miles){data-height=250}
```{r Table1_outOfDir}
base_df <- base_data[[which(base_csv_names=="avgStopOutofDirectionDist_vis")]]
build_df <- build_data[[which(build_csv_names=="avgStopOutofDirectionDist_vis")]]
df <- data.frame(base_df, build_df[,-1])
colnames(df) <- c("Tour_Purpose", BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME)
df$Tour_Purpose <- purpose_type_df$name[match(df$Tour_Purpose, purpose_type_df$code)]
#
t1 <- kable(df, format = "html", digits = 2, row.names = F, align = 'c', format.args = list(big.mark = ',')) %>%
  kable_styling("striped", full_width = F)

#t1 <- htmlTable(txtRound(df, 2),
#                align = "c|r",
#                rnames = F,
#                col.columns = c(rep("#E6E6F0", 1),
#                          rep("none", ncol(df) - 1)),
#                caption = "_______________________________________________________")

t1
```


TOD{data-navmenu="Trip Level"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Stop Departure**

Results of the Stop Departure Time Choice Model. The departure time of each stop on the tour is currently implemented as a Monte Carlo choice of time period from distributions generated from survey data.

The entire day is divided into 18 one-hour bins (The first bin includes 3:00 AM to 6:00 AM and the last bin includes 11:00 PM to 3:00 AM).

**Trip Departure**

Summarizes all trips by departure time period, including trips to and from intermediate stops and the tour primary destination.

Chart Column 1 {.tabset}
--------------------------------------------

### Stop & Trip Departure{data-height=650}
```{r stopDep}
base_df <- base_data[[which(base_csv_names=="stopTripDep_vis")]]
build_df <- build_data[[which(build_csv_names=="stopTripDep_vis")]]
colnames(base_df) <- c("timebin", "purpose", "freq_stop", "freq_trip")
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.purpose <- get_sharedData(data_df1 = base_df, data_df2 = build_df, run1_name = BASE_SCENARIO_NAME,
                             run2_name = BUILD_SCENARIO_NAME, x = "timebin", y = c("freq_stop", "freq_trip"), grp = "purpose")

p1 <- plotly_density_plotter(sd.purpose, index = "one", xlabel = "Stop Departure", percent = T, left_offset = 25,
                             tickvals = seq(1,48), ticktext = todBins, bottom_offset = 150, tickangle = 315, height = 400)
p2 <- plotly_density_plotter(sd.purpose, index = "two", xlabel = "Trip Departure", percent = T, left_offset = 25,
                             tickvals = seq(1,48), ticktext = todBins, bottom_offset = 150, tickangle = 315, height = 400)
#p3 <- datatable(sd.purpose$data())
bscols(widths=c(2,10),
  filter_select("Tour Purpose", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F),
  list(p1, p2)
  )
```

### Aggregate Stop & Trip Departure
```{r trip_tod_agg}
base_df <- base_data[[which(base_csv_names=="stopTripDep_vis")]]
colnames(base_df) <- c("id","purpose","freq_stop","freq_trip")
base_df$tod_agg <- cut(base_df$id, breaks = timePeriodBreaks, labels = timePeriods, right = FALSE)
base_df$tod_agg <- factor(base_df$tod_agg, levels = timePeriodOrder)
base_df <- base_df %>%
  group_by(purpose, tod_agg) %>%
  dplyr::summarise(freq_stop = sum(freq_stop), freq_trip = sum(freq_trip)) %>%
  ungroup()

build_df <- build_data[[which(build_csv_names=="stopTripDep_vis")]]
colnames(build_df) <- c("id","purpose","freq_stop","freq_trip")
build_df$tod_agg <- cut(build_df$id, breaks = timePeriodBreaks, labels = timePeriods, right = FALSE)
build_df$tod_agg <- factor(build_df$tod_agg, levels = timePeriodOrder)

build_df <- build_df %>%
  group_by(purpose, tod_agg) %>%
  dplyr::summarise(freq_stop = sum(freq_stop), freq_trip = sum(freq_trip)) %>%
  ungroup()
colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]

sd.purpose <- get_standardDF(data_df1=base_df, data_df2=build_df, x="tod_agg", y = c("freq_stop", "freq_trip"), grp = "purpose", shared = T)

p1 <- plotly_bar_plotter(data = sd.purpose, height = 350, xlabel = "Stop Departure", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T)
p2 <- plotly_bar_plotter(data = sd.purpose, height = 350, xlabel = "Trip Departure", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, index = 2)

bscols(widths=c(2,10),
  filter_select("Tour Purpose", "Select Tour Purpose", sd.purpose, ~grp_var,multiple=F),
  list(p1, p2)
  )


```



Trip Mode{data-navmenu="Trip Level"}
============================================



Chart Column 1 {data-width=125}
--------------------------------------------

###  {data-height=200}

***Trip Mode Choice***

The results of the Trip Mode Choice Model, which predicts the mode of each trip on the tour.

Distribution of trips by trip mode and tour mode, which constrains the availability of each trip mode and influences the utility of each available trip mode.

### Trip Mode Choice
```{r tripMode}
if(IS_BASE_SURVEY=="Yes"){
  base_df <- base_data[[which(base_csv_names=="tripModeProfile_vis_HTS")]]
}else{
  base_df <- base_data[[which(base_csv_names=="tripModeProfile_vis_HTS")]]
}
if(BUILD_SCENARIO_NAME == 'SEMCOG_HTS') {
  build_df <- build_data[[which(build_csv_names=="tripModeProfile_vis")]]
} else {
  build_df <- build_data[[which(build_csv_names=="tripModeProfile_vis")]]

}


colnames(build_df) <- colnames(base_df)

# change purpose names to standard format
base_df$purpose <- as.character(base_df$purpose)
build_df$purpose <- as.character(build_df$purpose)
base_df$purpose <- purpose_type_df$name[match(base_df$purpose, purpose_type_df$code)]
build_df$purpose <- purpose_type_df$name[match(build_df$purpose, purpose_type_df$code)]


sd.purpose <- get_standardDF(data_df1=base_df, data_df2=build_df, x="tripmode", y = c("value"), grp = "grp_var", shared = T)

p <- plotly_bar_plotter(data = sd.purpose, height = 700, xlabel = "Trip Mode", ylabel = "Percent", ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME), percent = T, tickvals = seq(1,length(tripMode)), ticktext = tripMode, bottom_offset = 75)

bscols(widths=c(12),
  list(filter_select("tripMode1", "Select Tour Purpose", sd.purpose, ~purpose,multiple=F),
       filter_select("tripMode1", "Select Tour Mode", sd.purpose, ~tourmode,multiple=F))
  )
```

Chart Column 2 {data-width=800}
--------------------------------------------
###
```{r tripMode2}
bscols(widths=c(12),
  list(p)
  )

```

Count vs Volume: All Day{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Comparison of observed counts and assigned volumes on each link with a counted volume, by assignment time period.


Chart Column 2{.tabset}
--------------------------------------------

### Count vs Volume - All Links{data-height=575}
```{r count_vol2}
ggplot(hnet[hnet$DY_CNT > 0,], aes(x = DY_CNT, y = DY_ASN)) + geom_point() +
	geom_abline(slope = 1, intercept = 0, color = "red") +
	scale_x_continuous("Daily Count", labels = comma) +
	scale_y_continuous("Daily Assignment", labels = comma)

```

### RMSE Statistics{data-height=575}
```{r rmse_vol1}

ggplot(vgsum, aes(x = vgidx, y = prmse_DAILY, color = "Model")) + geom_line(stat = "identity", size = 1.2) +
  geom_line(aes(y = limit/100, color = "Limit"), stat = "identity", size = 1.2) +
  scale_y_continuous("Percent RMSE", labels = percent) +
  scale_x_continuous("Volume Group", labels = function(x){return(vgsum[x + 1,"vg"])}, breaks = 0:nrow(vgsum)) + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```

### Assigned VMT Statistics{data-height=575}
```{r vmtcomp_dy}
  df = rbind.data.frame(cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$asnvmt_DAILY, "source" = "Model"), cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$obsvmt_DAILY, "source" = "Observed"))
df$VMT = as.numeric(df$VMT)

  ggplot(df, aes(x = FTYPE, y = VMT, fill = source)) + geom_bar(stat = "identity", position = "dodge") +
    scale_x_discrete("") +
    scale_color_discrete("Source") +
    scale_y_continuous("Vehicle Miles of Travel", labels = comma) 

```

Count vs Volume: AM{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Comparison of observed counts and assigned volumes on each link with a counted volume, by assignment time period.


Chart Column 2{.tabset}
--------------------------------------------

### AM Count vs Volume - All Links{data-height=575}
```{r count_vol6}
ggplot(hnet[hnet$AM_CNT > 0,], aes(x = AM_CNT, y = AM_ASN)) + geom_point() +
	geom_abline(slope = 1, intercept = 0, color = "red") +
	scale_x_continuous("AM Count", labels = comma) +
	scale_y_continuous("AM Assignment", labels = comma)
```

### RMSE Statistics{data-height=575}
```{r rmse_vol2}

ggplot(vgsum, aes(x = vgidx, y = prmse_AM, color = "Model")) + geom_line(stat = "identity", size = 1.2) +
  geom_line(aes(y = limit/100, color = "Limit"), stat = "identity", size = 1.2) +
  scale_y_continuous("Percent RMSE", labels = percent) +
  scale_x_continuous("Volume Group", labels = function(x){return(vgsum[x + 1,"vg"])}, breaks = 0:nrow(vgsum)) + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```

### Assigned VMT Statistics{data-height=575}
```{r vmtcomp_am}
  df = rbind.data.frame(cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$asnvmt_AM, "source" = "Model"), cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$obsvmt_AM, "source" = "Observed"))
df$VMT = as.numeric(df$VMT)

  ggplot(df, aes(x = FTYPE, y = VMT, fill = source)) + geom_bar(stat = "identity", position = "dodge") +
    scale_x_discrete("") +
    scale_color_discrete("Source") +
    scale_y_continuous("Vehicle Miles of Travel", labels = comma) 

```

Count vs Volume: MD{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Comparison of observed counts and assigned volumes on each link with a counted volume, by assignment time period.


Chart Column 2{.tabset}
--------------------------------------------

### MD Count vs Volume - All Links{data-height=575}
```{r count_vol8}
ggplot(hnet[hnet$MD_CNT > 0,], aes(x = MD_CNT, y = MD_ASN)) + geom_point() +
	geom_abline(slope = 1, intercept = 0, color = "red") +
	scale_x_continuous("Midday Count", labels = comma) +
	scale_y_continuous("Midday Assignment", labels = comma)
```

### RMSE Statistics{data-height=575}
```{r rmse_vol3}

ggplot(vgsum, aes(x = vgidx, y = prmse_MD, color = "Model")) + geom_line(stat = "identity", size = 1.2) +
  geom_line(aes(y = limit/100, color = "Limit"), stat = "identity", size = 1.2) +
  scale_y_continuous("Percent RMSE", labels = percent) +
  scale_x_continuous("Volume Group", labels = function(x){return(vgsum[x + 1,"vg"])}, breaks = 0:nrow(vgsum)) + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```

### Assigned VMT Statistics{data-height=575}
```{r vmtcomp_md}
  df = rbind.data.frame(cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$asnvmt_MD, "source" = "Model"), cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$obsvmt_MD, "source" = "Observed"))
df$VMT = as.numeric(df$VMT)

  ggplot(df, aes(x = FTYPE, y = VMT, fill = source)) + geom_bar(stat = "identity", position = "dodge") +
    scale_x_discrete("") +
    scale_color_discrete("Source") +
    scale_y_continuous("Vehicle Miles of Travel", labels = comma) 

```

Count vs Volume: PM{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Comparison of observed counts and assigned volumes on each link with a counted volume, by assignment time period.


Chart Column 2{.tabset}
--------------------------------------------

### PM Count vs Volume - All Links{data-height=575}
```{r count_vol10}
ggplot(hnet[hnet$PM_CNT > 0,], aes(x = PM_CNT, y = PM_ASN)) + geom_point() +
	geom_abline(slope = 1, intercept = 0, color = "red") +
	scale_x_continuous("PM Count", labels = comma) +
	scale_y_continuous("PM Assignment", labels = comma)
```

### RMSE Statistics{data-height=575}
```{r rmse_vol4}

ggplot(vgsum, aes(x = vgidx, y = prmse_PM, color = "Model")) + geom_line(stat = "identity", size = 1.2) +
  geom_line(aes(y = limit/100, color = "Limit"), stat = "identity", size = 1.2) +
  scale_y_continuous("Percent RMSE", labels = percent) +
  scale_x_continuous("Volume Group", labels = function(x){return(vgsum[x + 1,"vg"])}, breaks = 0:nrow(vgsum)) + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```

### Assigned VMT Statistics{data-height=575}
```{r vmtcomp_pm}
  df = rbind.data.frame(cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$asnvmt_PM, "source" = "Model"), cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$obsvmt_PM, "source" = "Observed"))
df$VMT = as.numeric(df$VMT)

  ggplot(df, aes(x = FTYPE, y = VMT, fill = source)) + geom_bar(stat = "identity", position = "dodge") +
    scale_x_discrete("") +
    scale_color_discrete("Source") +
    scale_y_continuous("Vehicle Miles of Travel", labels = comma) 

```

Count vs Volume: NT{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Comparison of observed counts and assigned volumes on each link with a counted volume, by assignment time period.


Chart Column 2{.tabset}
--------------------------------------------

### NT Count vs Volume - All Links{data-height=575}
```{r count_vol12}
ggplot(hnet[hnet$NT_CNT > 0,], aes(x = NT_CNT, y = NT_ASN)) + geom_point() +
	geom_abline(slope = 1, intercept = 0, color = "red") +
	scale_x_continuous("Nighttime Count", labels = comma) +
	scale_y_continuous("Nighttime Assignment", labels = comma)
```

### RMSE Statistics{data-height=575}
```{r rmse_vol5}

ggplot(vgsum, aes(x = vgidx, y = prmse_NT, color = "Model")) + geom_line(stat = "identity", size = 1.2) +
  geom_line(aes(y = limit/100, color = "Limit"), stat = "identity", size = 1.2) +
  scale_y_continuous("Percent RMSE", labels = percent) +
  scale_x_continuous("Volume Group", labels = function(x){return(vgsum[x + 1,"vg"])}, breaks = 0:nrow(vgsum)) + 
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1))

```

### Assigned VMT Statistics{data-height=575}
```{r vmtcomp_nt}
  df = rbind.data.frame(cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$asnvmt_NT, "source" = "Model"), cbind("FTYPE" = vmtcomp$FTYPE, "VMT" = vmtcomp$obsvmt_NT, "source" = "Observed"))
df$VMT = as.numeric(df$VMT)

  ggplot(df, aes(x = FTYPE, y = VMT, fill = source)) + geom_bar(stat = "identity", position = "dodge") +
    scale_x_discrete("") +
    scale_color_discrete("Source") +
    scale_y_continuous("Vehicle Miles of Travel", labels = comma) 

```

Total Assignment Summaries{data-navmenu="Assignment"}
============================================

Description {.sidebar data-width=175}
--------------------------------------------

********

**Link level count comparison**

Results of auto assignment.

Chart Column 2{.tabset}
--------------------------------------------

### Vehicle Miles of Travel{data-height=5750}
```{r asn_vmt}

build_df <- vmtsum[vmtsum$Period == "DAILY" & vmtsum$COUNTY != 10,]

build_df$CNAME = assign_county[build_df$COUNTY]

base_df = build_df
base_df$vmt = 0

#FTYPE,COUNTY,Period,nLinks,vmt

sd.asnvmt <- get_standardDF(data_df1 = base_df, data_df2 = build_df, x = "FTYPE", y = c("vmt"), grp = c("CNAME"), shared = TRUE)

p <- plotly_bar_plotter(data = sd.asnvmt, xlabel = "Facility Type", ylabel = "VMT", percent = T, height = 225, ynames = c(BASE_SCENARIO_NAME, BUILD_SCENARIO_NAME))

bscols(widths=c(3,9),
  list(
    filter_select("asnvmt_cname", "Select County", sd.asnvmt, ~grp_var, multiple=F)),
    p
)



```