Proyecciones RMNA 202526.qmd

---
title: "Proyecciones RMNA 2025-26 - GTRM Perú"
format: html-pdf
editor: visual
mainfont: Dosis
---

```{r setup, include = FALSE}
library(tidyverse)
library(stringr)
library(unhcrthemes)
library(scales)
 
# Source: Movements Report
data <- read.csv("https://raw.githubusercontent.com/IM-Peru/JNA_2024/main/Movement_Report.csv") 

data$Mes <- gsub(" ", "", data$Mes)

names(data) <- gsub("\\.", "", names(data))

# Two starting dates for two scenarios
fecha_1 <- "2023-12-31"
fecha_2 <- "2023-04-01"

# Stock for 2024
stock_2024 <- 1542004

data_1 <- data |>
  filter(Nacionalidad == "VEN") |> 
  mutate(Fecha = paste(Mes, "-01", sep="")) |>
  mutate(Fecha = as.Date(Fecha)) |>
  filter(Fecha >= fecha_1) |>
  select(Mes:"IrregularesTOTALSalidas") |>
  group_by(Nacionalidad) |>
  summarize(across(contains("egulare"), list(mean = ~mean(., na.rm = TRUE),
                                             min = ~min(., na.rm = TRUE),
                                             max = ~max(., na.rm = TRUE),
                                             median = ~median(., na.rm = TRUE)), .names = "{.col}_{.fn}")) |>
  pivot_longer(-1, names_to = c("Tipo", "Calc"), names_sep = "_", values_to = "Tot") |>
  mutate(Tot = round(Tot, 0)) |>
  ungroup() |>
  filter(!grepl('DTM', Tipo)) |>
  pivot_wider(names_from = Calc, values_from = Tot) |>
  mutate(Flujo = if_else(str_detect(Tipo, "Regulares"), "Regulares", "Irregulares")) |>
  mutate(Punto = if_else(str_detect(Tipo, "Tumbes"), "Tumbes", 
                         if_else(str_detect(Tipo, "Tacna"), "Tacna", 
                                 if_else(str_detect(Tipo, "Puno"), "Puno", 
                                         if_else(str_detect(Tipo, "Madre"), "MdD", 
                                                 if_else(str_detect(Tipo, "Aeropuerto"), "Aeropuerto", 
                                                         if_else(str_detect(Tipo, "otros"), "Otros", "Total"))))))) |>
  mutate(Direct = if_else(str_detect(Tipo, "Entradas"), "Entradas", "Salidas")) |>
  select(Nacionalidad, Punto, Flujo, Direct, mean, min, max, median)
  


data_1 <- lapply(data_1, function(x) {
  x[is.infinite(x)] <- NA
  x[is.nan(x)] <- NA
  return(x)
})

data_1 <- as.data.frame(data_1)
data_1 <- data_1[complete.cases(data_1), ]


# Cálculo escenarios desde Fecha 2

data_2 <- data |>
  mutate(Fecha = paste(Mes, "-01", sep="")) |>
  mutate(Fecha = as.Date(Fecha)) |>
  filter(Fecha >= fecha_2) |>
  select(Mes:"IrregularesTOTALSalidas") |>
  group_by(Nacionalidad) |>
  summarize(across(contains("egulare"), list(mean = ~mean(., na.rm = TRUE),
                                             min = ~min(., na.rm = TRUE),
                                             max = ~max(., na.rm = TRUE),
                                             median = ~median(., na.rm = TRUE)), .names = "{.col}_{.fn}")) |>
  pivot_longer(-1, names_to = c("Tipo", "Calc"), names_sep = "_", values_to = "Tot") |>
  mutate(Tot = round(Tot, 0)) |>
  ungroup() |>
  filter(!grepl('DTM', Tipo)) |>
  pivot_wider(names_from = Calc, values_from = Tot) |>
  mutate(Flujo = if_else(str_detect(Tipo, "Regulares"), "Regulares", "Irregulares")) |>
  mutate(Punto = if_else(str_detect(Tipo, "Tumbes"), "Tumbes", 
                         if_else(str_detect(Tipo, "Tacna"), "Tacna", 
                                 if_else(str_detect(Tipo, "Puno"), "Puno", 
                                         if_else(str_detect(Tipo, "Madre"), "MdD", 
                                                 if_else(str_detect(Tipo, "Aeropuerto"), "Aeropuerto", 
                                                         if_else(str_detect(Tipo, "otros"), "Otros", "Total"))))))) |>
  mutate(Direct = if_else(str_detect(Tipo, "Entradas"), "Entradas", "Salidas")) |>
  select(Nacionalidad, Punto, Flujo, Direct, mean, min, max, median) |>
  filter(Nacionalidad == "VEN")


data_2 <- lapply(data_2, function(x) {
  x[is.infinite(x)] <- NA
  x[is.nan(x)] <- NA
  return(x)
})

data_2 <- as.data.frame(data_2)
data_2 <- data_2[complete.cases(data_2), ]


```

## Saldo Escenario 1: Flujos 2024

Fecha de referencia: desde `r fecha_1` a 2024-04-31

::: columns
::: {.column width="60%"}
```{r, echo = FALSE, warning = FALSE}

saldo <- function(direction, var){
    result = data_1 |> 
    filter(Punto != "Total") |> 
    filter(Direct == direction) |>
    select(Direct, {{ var }}) |> 
    group_by(Direct) |> 
    summarize(value = sum({{ var }}, na.rm = TRUE))
    return(result$value)
}

entries_mean_1 <- saldo("Entradas", mean) / 30
entries_min_1 <- saldo("Entradas", min) / 30
entries_max_1 <- saldo("Entradas", max) / 30
entries_max_1 <- saldo("Entradas", max) / 30
entries_median_1 <- saldo("Entradas", median) / 30

exits_mean_1 <- saldo("Salidas", mean) / 30
exits_min_1 <- saldo("Salidas", min) / 30
exits_max_1 <- saldo("Salidas", max) / 30
exits_median_1 <- saldo("Salidas", median) / 30


# entries <- list(mean_1 = entries_mean_1, min_1 = entries_min_1, max_1 = entries_max_1)
# exits <- list(mean_1 = exits_mean_1, min_1 = exits_min_1, max_1 = exits_max_1)

result_matrix <- matrix(nrow = 4, ncol = 4)
rownames(result_matrix) <- c("entries_mean_1", "entries_min_1", "entries_max_1", "entries_median_1")
colnames(result_matrix) <- c("exits_mean_1", "exits_min_1", "exits_max_1", "exits_median_1")

result_matrix["entries_mean_1", "exits_mean_1"] <- entries_mean_1 - exits_mean_1
result_matrix["entries_mean_1", "exits_min_1"]  <- entries_mean_1 - exits_min_1
result_matrix["entries_mean_1", "exits_max_1"]  <- entries_mean_1 - exits_max_1
result_matrix["entries_mean_1", "exits_median_1"]  <- entries_mean_1 - exits_median_1

result_matrix["entries_min_1", "exits_mean_1"]  <- entries_min_1 - exits_mean_1
result_matrix["entries_min_1", "exits_min_1"]   <- entries_min_1 - exits_min_1
result_matrix["entries_min_1", "exits_max_1"]   <- entries_min_1 - exits_max_1
result_matrix["entries_min_1", "exits_median_1"]   <- entries_min_1 - exits_median_1


result_matrix["entries_max_1", "exits_mean_1"]  <- entries_max_1 - exits_mean_1
result_matrix["entries_max_1", "exits_min_1"]   <- entries_max_1 - exits_min_1
result_matrix["entries_max_1", "exits_max_1"]   <- entries_max_1 - exits_max_1
result_matrix["entries_max_1", "exits_median_1"]   <- entries_max_1 - exits_median_1

result_matrix["entries_median_1", "exits_mean_1"]  <- entries_median_1 - exits_mean_1
result_matrix["entries_median_1", "exits_min_1"]   <- entries_median_1 - exits_min_1
result_matrix["entries_median_1", "exits_max_1"]   <- entries_median_1 - exits_max_1
result_matrix["entries_median_1", "exits_median_1"]   <- entries_median_1 - exits_median_1



saldo_diario <- as.data.frame(as.table(result_matrix)) |> 
  mutate(across(Freq,\(x) round(x, 0))) |> 
  rename(Entries = Var1) |> 
  rename(Exits = Var2)|>
  mutate(Entries = if_else(Entries == "entries_max_1", "Máximo Ingresos", 
                           if_else(Entries == "entries_min_1", "Mínimo Ingresos", 
                                   if_else(Entries == "entries_mean_1", "Promedio Ingresos",
                                   "Ingresos Medianos")))) |> 
  mutate(Exits = if_else(Exits == "exits_max_1", "Máximo Salidas", 
                         if_else(Exits == "exits_min_1", "Mínimo Salidas", 
                                  if_else(Exits == "exits_mean_1", "Promedio Salidas", 
                                 "Salidas Medianas"))))


#print(result_matrix)

ggplot(
  saldo_diario |> 
    mutate(Entries = fct_relevel(Entries, c("Máximo Ingresos", "Mínimo Ingresos", "Ingresos Medianos", "Promedio Ingresos")),
                Exits = fct_relevel(Exits, c("Promedio Salidas", "Salidas Medianas", "Mínimo Salidas", "Máximo Salidas"))),
  aes(
    x = Exits,
    y = Entries
  )
) +
  geom_tile(aes(
    fill = Freq
  ),
  color = "white",
  lwd = .5,
  linetype = 1
  ) +
  labs(
    title = "Saldo diario 2025-26"
  ) +
  scale_x_discrete(
    labels = scales::label_wrap(3),
    position = "top"
  ) +
  scale_y_discrete(
    labels = scales::label_wrap(3)
  ) +
  scale_fill_stepsn(
    colors = c("#75B5E4", "#0088CC", "#00649F", "#004469"),
    n.break = 3,
    name = "Saldo diario"
  ) +
  coord_fixed() +
   geom_text(aes(label = Freq), colour = "white", size = 3
  ) +
  theme_unhcr(
    font_size = 13,
    grid = FALSE,
    axis = FALSE,
    axis_title = FALSE,
    legend_title = TRUE
  )



```

:::

::: {.column width="20%"}
 \
**Rangos:**  \
 \

**`r round(sum(entries_max_1 - exits_min_1), 0)`** <br/> Saldo máximo <br/> (`r format(round(sum(entries_max_1 - exits_min_1), 0)*365, big.mark = ",")` anual)


 \

**`r round(sum(entries_min_1 - exits_max_1), 0)`** <br/> Saldo mínimo <br/> (`r format(round(sum(entries_min_1 - exits_max_1), 0)*365, big.mark = ",")` anual)

 \

:::

::: {.column width="20%"}
 \
**Status quo:**  \
 \

**`r round(sum(entries_mean_1 - exits_mean_1), 0)`** <br/> Saldo promedio <br/> (`r format(round(sum(entries_mean_1 - exits_mean_1), 0)*365, big.mark = ",")` anual)


 \

**`r round(sum(entries_median_1 - exits_median_1), 0)`** <br/> Saldo mediano <br/> (`r format(round(sum(entries_median_1 - exits_median_1), 0)*365, big.mark = ",")` anual)
:::

:::

La tabla de calor muestra las posibles combinaciones de los saldos según los flujos observados en los meses de 2024, calculados utilizando los valores máximos, mínimos, promedio y mediano. Por ejemplo, la primera celda arriba la izquierda enseña el saldo diario restando el promedio de salidas al promedio de entradas durante el periodo señalado.

En este sentido se consideran que los rangos de proyecciones se pueden categorizar entre (1) un **saldo máximo**, donde se plantea un máximo de ingresos y un mínimo de salidas al país y (2) un **saldo mínimo**, en el cual se considera un mínimo de ingresos y un máximo de salidas al país . Entre ellos, se puede estimar un **resultado status quo**, representado por las entradas y salidas promedio o medianas.

```{r, echo = FALSE, warning = FALSE, fig.height = 4.5, fig.width = 8}

year <- c(2024, 2025, 2026)
mean_1 <- c(stock_2024, stock_2024 + (sum(entries_mean_1 - exits_mean_1) * 365), stock_2024 + ((sum(entries_mean_1 - exits_mean_1) * 365))*2)
min_1 <- c(stock_2024, stock_2024 + (sum(entries_min_1 - exits_max_1) * 365), stock_2024 + ((sum(entries_min_1 - exits_max_1) * 365))*2)
max_1 <- c(stock_2024, stock_2024 + (sum(entries_max_1 - exits_min_1) * 365), stock_2024 + ((sum(entries_max_1 - exits_min_1) * 365))*2)
median_1 <- c(stock_2024, stock_2024 + (sum(entries_median_1 - exits_median_1) * 365), stock_2024 + ((sum(entries_median_1 - exits_median_1) * 365))*2)


stock_proj_1 <- data.frame(year, mean_1, min_1, max_1, median_1) |> 
  pivot_longer(-1, names_to = "type", values_to = "stock") |> 
  mutate(type = str_sub(type, end = -3)) |> 
  mutate(year = as.character(year)) |>
  mutate(type = if_else(type == "mean", "Promedio", 
                        if_else(type == "min", "Mínimo", 
                               if_else(type == "max", "Máximo", 
                                "Mediana"))))



# Plot
ggplot(stock_proj_1 |> 
         mutate(type = fct_relevel(type, c("Máximo", "Mínimo", "Promedio", "Mediana")),
                year = fct_relevel(year, c("2026", "2025", "2024")))) +
  geom_col(aes(
    x = type,
    y = stock,
    fill = fct_rev(year)
  ),
  position = position_dodge(width = 0.7),
  width = 0.6
  ) +
  geom_text(aes(
    x = type,
    y = stock,
    group = fct_rev(year),
    label = round(stock / 1e6, 3)
  ),
  position = position_dodge(width = 0.7),
  vjust = -1,
  size = 8 / .pt
  ) +
  scale_fill_unhcr_d(
    palette = "pal_unhcr",
    nmax = 3,
    order = c(3, 2, 1)
  ) +
  labs(
    title = "Proyecciones 2025-26",
    subtitle = "En millones de personas"
  ) +
  # scale_x_continuous(breaks = pretty_breaks(n = 4)) +
  # scale_y_continuous(expand = expansion(c(0, 0.1))) +
  theme_unhcr(
    grid = FALSE,
    axis = "x",
    axis_title = FALSE,
    axis_text = "x"
  ) +
  guides(fill = guide_legend(nrow = 1, byrow = TRUE)) +
  ylim(0, max(stock_proj_1$stock)*1.1)


```

## Saldo Escenario 2: Últimos 12 meses

Fecha de referencia: desde `r fecha_2` a 2024-04-31

::: columns
::: {.column width="60%"}
```{r, echo = FALSE, warning = FALSE}

saldo <- function(direction, var){
  result = data_2 |> 
    filter(Punto != "Total") |> 
    filter(Direct == direction) |>
    select(Direct, {{ var }}) |> 
    group_by(Direct) |> 
    summarize(value = sum({{ var }}, na.rm = TRUE))
    return(result$value)
}

entries_mean_2 <- saldo("Entradas", mean) / 30
entries_min_2 <- saldo("Entradas", min) / 30
entries_max_2 <- saldo("Entradas", max) / 30
entries_median_2 <- saldo("Entradas", median) / 30


exits_mean_2 <- saldo("Salidas", mean) / 30
exits_min_2 <- saldo("Salidas", min) / 30
exits_max_2 <- saldo("Salidas", max) / 30
exits_median_2 <- saldo("Salidas", median) / 30

# entries <- list(mean_2 = entries_mean_2, min_2 = entries_min_2, max_2 = entries_max_2)
# exits <- list(mean_2 = exits_mean_2, min_2 = exits_min_2, max_2 = exits_max_2)

result_matrix <- matrix(nrow = 4, ncol = 4)
rownames(result_matrix) <- c("entries_mean_2", "entries_min_2", "entries_max_2", "entries_median_2")
colnames(result_matrix) <- c("exits_mean_2", "exits_min_2", "exits_max_2", "exits_median_2")

result_matrix["entries_mean_2", "exits_mean_2"] <- entries_mean_2 - exits_mean_2
result_matrix["entries_mean_2", "exits_min_2"]  <- entries_mean_2 - exits_min_2
result_matrix["entries_mean_2", "exits_max_2"]  <- entries_mean_2 - exits_max_2
result_matrix["entries_mean_2", "exits_median_2"]  <- entries_mean_2 - exits_median_2


result_matrix["entries_min_2", "exits_mean_2"]  <- entries_min_2 - exits_mean_2
result_matrix["entries_min_2", "exits_min_2"]   <- entries_min_2 - exits_min_2
result_matrix["entries_min_2", "exits_max_2"]   <- entries_min_2 - exits_max_2
result_matrix["entries_min_2", "exits_median_2"]   <- entries_min_2 - exits_median_2

result_matrix["entries_max_2", "exits_mean_2"]  <- entries_max_2 - exits_mean_2
result_matrix["entries_max_2", "exits_min_2"]   <- entries_max_2 - exits_min_2
result_matrix["entries_max_2", "exits_max_2"]   <- entries_max_2 - exits_max_2
result_matrix["entries_max_2", "exits_median_2"]   <- entries_max_2 - exits_median_2

result_matrix["entries_median_2", "exits_mean_2"]  <- entries_median_2 - exits_mean_2
result_matrix["entries_median_2", "exits_min_2"]   <- entries_median_2 - exits_min_2
result_matrix["entries_median_2", "exits_max_2"]   <- entries_median_2 - exits_max_2
result_matrix["entries_median_2", "exits_median_2"]   <- entries_median_2 - exits_median_2


saldo_diario <- as.data.frame(as.table(result_matrix)) |> 
  mutate(across(Freq,\(x) round(x, 0))) |> 
  rename(Entries = Var1) |> 
  rename(Exits = Var2)|>
  mutate(Entries = if_else(Entries == "entries_max_2", "Máximo Ingresos", 
                           if_else(Entries == "entries_min_2", "Mínimo Ingresos", 
                                    if_else(Entries == "entries_mean_2", "Promedio Ingresos",
                                   "Ingresos Medianos")))) |> 
  mutate(Exits = if_else(Exits == "exits_max_2", "Máximo Salidas", 
                         if_else(Exits == "exits_min_2", "Mínimo Salidas", 
                                 if_else(Exits == "exits_mean_2", "Promedio Salidas", 
                                 "Salidas Medianas"))))

#print(result_matrix)

ggplot(
  saldo_diario |> 
    mutate(Entries = fct_relevel(Entries, c("Máximo Ingresos", "Mínimo Ingresos", "Ingresos Medianos", "Promedio Ingresos")),
                Exits = fct_relevel(Exits, c("Promedio Salidas", "Mínimo Salidas", "Máximo Salidas", "Salidas Medianas"))),
  aes(
    x = Exits,
    y = Entries
  )
) +
  geom_tile(aes(
    fill = Freq
  ),
  color = "white",
  lwd = .5,
  linetype = 1
  ) +
  labs(
    title = "Saldo diario 2025-26"
  ) +
  scale_x_discrete(
    labels = scales::label_wrap(3),
    position = "top"
  ) +
  scale_y_discrete(
    labels = scales::label_wrap(3)
  ) +
  scale_fill_stepsn(
    colors = c("#75B5E4", "#0088CC", "#00649F", "#004469"),
    n.break = 3,
    name = "Saldo diario"
  ) +
  coord_fixed() +
   geom_text(aes(label = Freq), colour = "white", size = 3
  ) +
  theme_unhcr(
    font_size = 13,
    grid = FALSE,
    axis = FALSE,
    axis_title = FALSE,
    legend_title = TRUE
  )



```

:::

::: {.column width="20%"}
 \
**Rangos:**  \
 \

**`r round(sum(entries_max_2 - exits_min_2), 0)`** <br/> Saldo máximo <br/> (`r format(round(sum(entries_max_2 - exits_min_2), 0)*365, big.mark = ",")` anual)


 \

**`r round(sum(entries_min_2 - exits_max_2), 0)`** <br/> Saldo mínimo <br/> (`r format(round(sum(entries_min_2 - exits_max_2), 0)*365, big.mark = ",")` anual)

 \

:::

::: {.column width="20%"}
 \
**Status quo:**  \
 \

**`r round(sum(entries_mean_2 - exits_mean_2), 0)`** <br/> Saldo promedio <br/> (`r format(round(sum(entries_mean_2 - exits_mean_2), 0)*365, big.mark = ",")` anual)


 \

**`r round(sum(entries_median_2 - exits_median_2), 0)`** <br/> Saldo mediano <br/> (`r format(round(sum(entries_median_2 - exits_median_2), 0)*365, big.mark = ",")` anual)
:::

:::

La tabla de calor muestra las posibles combinaciones de los saldos según los flujos observados en los meses de 2024, calculados utilizando los valores máximos, mínimos, promedio y mediano. Por ejemplo, la primera celda arriba la izquierda enseña el saldo diario restando el promedio de salidas al promedio de entradas durante el periodo señalado.

En este sentido se consideran que los rangos de proyecciones se pueden categorizar entre (1) un **saldo máximo**, donde se plantea un máximo de ingresos y un mínimo de salidas al país y (2) un **saldo mínimo**, en el cual se considera un mínimo de ingresos y un máximo de salidas al país . Entre ellos, se puede estimar un **resultado status quo**, representado por las entradas y salidas promedio o medianas.
```{r, echo = FALSE, warning = FALSE, fig.height = 4.5, fig.width = 8}

year <- c(2024, 2025, 2026)
mean_2 <- c(stock_2024, stock_2024 + (sum(entries_mean_2 - exits_mean_2) * 365), stock_2024 + ((sum(entries_mean_2 - exits_mean_2) * 365))*2)
min_2 <- c(stock_2024, stock_2024 + (sum(entries_min_2 - exits_max_2) * 365), stock_2024 + ((sum(entries_min_2 - exits_max_2) * 365))*2)
max_2 <- c(stock_2024, stock_2024 + (sum(entries_max_2 - exits_min_2) * 365), stock_2024 + ((sum(entries_max_2 - exits_min_2) * 365))*2)
median_2 <- c(stock_2024, stock_2024 + (sum(entries_median_2 - exits_median_2) * 365), stock_2024 + ((sum(entries_median_2 - exits_median_2) * 365))*2)

stock_proj_2 <- data.frame(year, mean_2, min_2, max_2, median_2) |> 
  pivot_longer(-1, names_to = "type", values_to = "stock") |> 
  mutate(type = str_sub(type, end = -3)) |> 
  mutate(year = as.character(year)) |>
   mutate(type = if_else(type == "mean", "Promedio", 
                         if_else(type == "min", "Mínimo", 
                                 if_else(type == "max", "Máximo",
                                 "Mediana"))))



# Plot
ggplot(stock_proj_2 |> 
         mutate(type = fct_relevel(type, c("Máximo", "Mínimo", "Promedio","Mediana")),
                year = fct_relevel(year, c("2026", "2025", "2024")))) +
  geom_col(aes(
    x = type,
    y = stock,
    fill = fct_rev(year)
  ),
  position = position_dodge(width = 0.7),
  width = 0.6
  ) +
  geom_text(aes(
    x = type,
    y = stock,
    group = fct_rev(year),
    label = round(stock / 1e6, 3)
  ),
  position = position_dodge(width = 0.7),
  vjust = -1,
  size = 8 / .pt
  ) +
  scale_fill_unhcr_d(
    palette = "pal_unhcr",
    nmax = 3,
    order = c(3, 2, 1)
  ) +
  labs(
    title = "Proyecciones 2025-26",
    subtitle = "En millones de personas"
  ) +
  # scale_x_continuous(breaks = pretty_breaks(n = 4)) +
  # scale_y_continuous(expand = expansion(c(0, 0.1))) +
  theme_unhcr(
    grid = FALSE,
    axis = "x",
    axis_title = FALSE,
    axis_text = "x"
  ) +
  guides(fill = guide_legend(nrow = 1, byrow = TRUE)) +
  ylim(0, max(stock_proj_2$stock)*1.1)


```


<br/>


## Propuesta de Proyecciones 2025-26 - Equipo IM

Considerando estos elementos, el equipo de Manejo de Información propone las proyecciones siguientes:

::: {.column width="30%"}

<center>

**`r round(sum(entries_mean_2 - exits_mean_2), 0)`** <br/> Saldo proyectado <br/> (`r format(round(sum(entries_mean_2 - exits_mean_2), 0)*365, big.mark = ",")` anual)

</center>

:::

::: {.column width="30%"}

<center>

**`r format(round(stock_2024 + (sum(entries_mean_2 - exits_mean_2) * 365), 0), big.mark = ",")`** <br/> Stock 2025 <br/> 

</center>

:::

::: {.column width="30%"}

<center>

**`r format(round(stock_2024 + ((sum(entries_mean_2 - exits_mean_2) * 365)) * 2, 0), big.mark = ",")`** <br/> Stock 2026 <br/> 

</center>

:::


Esta proyección utiliza el saldo promedio de ingresos y salidas observados entre abril 2023 y abril 2024 y presupone un stock constante en 2024 de 1.54M de personas venezolanas. Los flujos bajo esta proyección serían los siguientes:



::: {.column width="49%" }

<center>

**`r format(round(sum(entries_mean_2), 0), big.mark = ",")`** <br>
Entradas diarias <br>
(`r format(round(sum(entries_mean_2), 0)*365, big.mark = ",")` anual)

</center>

::: 

::: {.column width="49%" }

<center>

**`r format(round(sum(exits_mean_2), 0), big.mark = ",")`** <br>
Salidas diarias <br>
(`r format(round(sum(exits_mean_2), 0)*365, big.mark = ",")` anual)

</center>

::: 

```{r, echo = FALSE, warning = FALSE, message = FALSE}

# Utilizando Promedio de Escenario 2
data_proyecciones <- data_2 |> 
  select(Punto, Direct, mean) |> 
  group_by(Punto, Direct) |> 
  summarize(mean = sum(mean, na.rm = TRUE)) |> 
  filter(Punto != "Total") |> 
  mutate(mean = round(mean / 30 * 365, 0)) |> 
  pivot_wider(names_from = Direct, values_from = mean) |> 
  arrange(Punto) |> 
  mutate(Entradas = format(Entradas, big.mark = ",")) |> 
  mutate(Salidas = format(Salidas, big.mark = ",")) |> 
  rename("Entradas anuales" = Entradas) |> 
  rename("Salidas anuales" = Salidas)


entries_mean_2 <- saldo("Entradas", mean) / 30


library(kableExtra)
kable(data_proyecciones, align=rep('r', 5), format = "html")
```


Aplicando esta proyección al stock de población venezolana en Perú en el tiempo, se obtendría lo siguiente:

```{r, echo = FALSE, warning = FALSE, message = FALSE, fig.height = 3.5, fig.width = 8}

stock_2016 <- 8170
stock_2017 <- 66858
stock_2018 <- 691565
stock_2019 <- 863613
stock_2020 <- 1043460
stock_2021 <- 1286464
stock_2022 <- 1493645
stock_2023 <- 1542004
stock_2024 <- 1542004
stock_2025 <- round(stock_2024 + (sum(entries_mean_2 - exits_mean_2) * 365), 0)
stock_2026 <- round(stock_2024 + ((sum(entries_mean_2 - exits_mean_2) * 365)*2), 0)

Stock <- c(stock_2016, stock_2017, stock_2018, stock_2019, stock_2020, stock_2021, stock_2022, stock_2023, stock_2024, stock_2025, stock_2026)
Stock_year <- c(2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026)

Stock <- data.frame(Stock_year, Stock) |> 
  mutate(Stock_year = as.character(Stock_year)) |> 
  select(Stock_year, Stock) 


# Plot
ggplot(Stock) +
  geom_col(aes(
    x = Stock_year,
    y = Stock
  ),
  fill = "#00649F",
  width = 0.8
  ) +
  geom_text(aes(
    x = Stock_year,
    y = Stock,
    label = paste(round(Stock / 1e6, 3), "M", sep = "")
  ),
  vjust = -1, size = 8 / .pt
  ) +
  labs(
    title = "Evolución de total poblacional en el tiempo"
  ) +
  scale_y_continuous(expand = expansion(c(0, 0.1))) +
  theme_unhcr(
    grid = FALSE,
    axis = "x",
    axis_title = FALSE,
    axis_text = "x"
  )

```