S7_O2_plots_new.Rmd

---
title: "STEPS Study: O2 Exploratory Plots"
author: "C. Konstantinou and X. Andrianou @ Water and Health Laboratory, CII-CUT"
date: "`r format(Sys.time(), '%d %B, %Y')`"
output:
  rmdformats::downcute:
    self_contained: true
    thumbnails: true
    lightbox: true
    gallery: false
    highlight: tango
editor_options: 
  chunk_output_type: console
---



```{r packages,  echo=FALSE, include=FALSE}
rm(list=ls())


ipak <- function(pkg){
  new.pkg <- pkg[!(pkg %in% installed.packages()[, "Package"])]
  if (length(new.pkg))
    install.packages(new.pkg, dependencies = TRUE,
                     repos = "http://cran.us.r-project.org")
  sapply(pkg, require, character.only = TRUE)
}

# usage
packages <- c("tidyverse","lubridate","officer", "rvg","glue", "patchwork","janitor",
              "tableone","dplyr","lme4","readxl","scales","readr", "lubridate", 
              "patchwork")

ipak(packages)
```



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

# read questionnaire_sensor files
questPA_1 <- readRDS("produced_data/questPA_1.rds", refhook = NULL)
questMCF_1 <- readRDS("produced_data/questMCF_1.rds", refhook = NULL)

dataset_final_regression <- readRDS("produced_data/dataset_final_regression.rds", refhook = NULL)
dataset_final_regression_M <- readRDS("produced_data/dataset_final_regression_M.rds", refhook = NULL)

dataset_final_regression_all <- readRDS("produced_data/dataset_final_regression_all.rds", refhook = NULL)
dataset_final_regression_all_M <- readRDS("produced_data/dataset_final_regression_all_M.rds", refhook = NULL)

dataset_final_regression_probandnorm <- readRDS("produced_data/dataset_final_regression_probandnorm.rds") %>% 
   mutate(class_type = as.factor(class_type))
dataset_final_regression_M_probandnorm <- readRDS("produced_data/dataset_final_regression_M_probandnorm.rds")
```


# Main text 

## Raw data without dates including classes hours (7:45-13:05) and time before and after classes hours (7:00 - 07:45 and 13:05 - 15:00) - all classrooms

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

# samplingtime" used for plots in S7 is dependent on S5. So, if S5 is re-run, then the date changes and date needs to be replaced here e.g. use the search and replace all function on top left side 

questPA_1 %>%
  group_by(name) %>% 
  filter(period_new !="no_school") %>%
  filter(final_value<quantile(final_value, probs=0.99, na.rm=TRUE)) %>%
  summarise(max=max(final_value, na.rm = T))

PA_params_plot <- questPA_1 %>%
  group_by(name) %>% 
  filter(period_new !="no_school") %>%
  filter(final_value<quantile(final_value, probs=0.99, na.rm=TRUE)) %>% 
  mutate(label_for_plot=case_when(name=="pm1_0"~"PM1 (ug/m3)",
                                  name=="pm2_5"~"PM2.5 (ug/m3)",
                                  name=="pm10_0"~"PM10 (ug/m3)")) %>%
  ggplot() +
   geom_point(aes(x=samplingtime, y=final_value, colour=indoor_outdoor),
              position=position_jitter(h=0.1, w=0.1),
             shape = 20, alpha = 0.1, size = 0.2) +
   geom_smooth(aes(x=samplingtime, y=final_value, colour=indoor_outdoor), size = 2, method="loess", span=0.1,se = F) +
  scale_color_manual(values = RColorBrewer::brewer.pal(n = 6, "Set2")[1:2])+
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   annotate("rect", 
            xmin=as.POSIXct("2020-01-01 09:05:00"), xmax=as.POSIXct("2020-01-01 09:25:00"),
           ymin=min(questPA_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
   annotate("rect", 
            xmin=as.POSIXct("2020-01-01 10:45:00"), xmax=as.POSIXct("2020-01-01 10:55:00"),
            ymin=min(questPA_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
      annotate("rect", 
            xmin=as.POSIXct("2020-01-01 12:15:00"), xmax=as.POSIXct("2020-01-01 12:25:00"),
            ymin=min(questPA_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
    #annotate classes starting time 07:45
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 07:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[3]) + 
    #annotate classes stopping time before classes begin13:05
  geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 13:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[3]) +
    #annotate school opening time 07:00 -  cleaning staff, teachers arrive
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 07:00:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4]) +
    #annotate school closing time 15:00 -  cleaning staff leave
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 15:00:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4]) +
    labs(y= NULL, x = NULL) +
  facet_wrap(~label_for_plot, ncol = 1, scales="free_y") +
   theme_bw() +
   theme(legend.position = "bottom",
         legend.title = element_blank())
  
       # only school, before and after - 7:00-15:00
   

questMCF_1 %>%
  group_by(name) %>% 
  filter(period_new !="no_school") %>%
  filter(final_value<quantile(final_value, probs=0.99, na.rm=TRUE)) %>%
  summarise(max=max(final_value, na.rm = T))

MCF_params_plot <- questMCF_1 %>%
  group_by(name) %>% 
  filter(period_new !="no_school") %>% 
  mutate(label_for_plot=case_when(name=="b_voc"~"VOCs (ppb)",
                                  name=="carbon_dioxide"~"CO2 (ppm)",
                                  name=="humidity"~"Humidity (%)",
                                  name=="temperature"~"Temperature (C)")) %>%
  filter(final_value<quantile(final_value, probs=0.99, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(x=samplingtime, y=final_value, colour=indoor_outdoor),
              position=position_jitter(h=0.1, w=0.1),
             shape = 20, alpha = 0.1, size = 0.2) +
   geom_smooth(aes(x=samplingtime, y=final_value, colour=indoor_outdoor), size = 2, method="loess", span=0.1,se = F) +
  scale_color_manual(values = RColorBrewer::brewer.pal(n = 6, "Set2")[1:2])+
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   annotate("rect", 
            xmin=as.POSIXct("2020-01-01 09:05:00"), xmax=as.POSIXct("2020-01-01 09:25:00"),
           ymin=min(questMCF_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
   annotate("rect", 
            xmin=as.POSIXct("2020-01-01 10:45:00"), xmax=as.POSIXct("2020-01-01 10:55:00"),
            ymin=min(questMCF_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
      annotate("rect", 
            xmin=as.POSIXct("2020-01-01 12:15:00"), xmax=as.POSIXct("2020-01-01 12:25:00"),
            ymin=min(questMCF_1$final_value, na.rm = T), ymax=Inf, alpha=0.1, fill="yellow") +
    #annotate classes starting time 07:45
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 07:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[3]) +
    #annotate classes stopping time before classes begin13:05
  geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 13:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[3]) +
    #annotate school opening time 07:00 -  cleaning staff, teachers arrive
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 07:00:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4]) +
    #annotate school closing time 15:00 -  cleaning staff leave
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 15:00:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4]) +
    labs(y= NULL, x = NULL) +
  facet_wrap(~label_for_plot, ncol = 1, scales="free_y") +
   theme_bw() +
   theme(legend.position = "bottom",
         legend.title = element_blank())
  
fig1 <- PA_params_plot + MCF_params_plot  + plot_layout(guides = "collect") & theme(legend.position = 'bottom')

fig1

ggsave(filename = "plots/Figure1.png", plot=fig1, dpi=300)
```

# Supplementary 

## Period means for all schools - comparison of normal and problematic classrooms

Periods: 

+ 1-10 = school periods from classes start time (7:45) to classes end time (13:05) e.g. 1 = 07:45 - 08:25
+ bef_classes = period from 7:00-7:45 
+ bef_classes = period from 7:00-7:45 
+ after_classes = period from 13:05-15:00 


+ 95th -ile

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

 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm10_0") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "PM10 (ug/m3)")

 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm1_0") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "PM1 (ug/m3)")

 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm2_5") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "PM2.5 (ug/m3)")

 # only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="temperature") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "Temperature (C)")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="humidity") %>%
     filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "Rel. Humidity")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="carbon_dioxide") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "CO2 (ppm)")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="b_voc") %>%
     filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,colour=class_type), 
              position=position_jitter(h=0.1, w=0.1)) + 
   theme_minimal() +
    labs(y= "VOCs (ppb)")



 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm10_0") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "PM10 (ug/m3)")

 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm1_0") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "PM1 (ug/m3)")

 # only school hours 
dataset_final_regression_probandnorm %>%
   filter(name=="pm2_5") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "PM2.5 (ug/m3)")

 # only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="temperature") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "Temperature (C)")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="humidity") %>%
     filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "Rel. Humidity")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="carbon_dioxide") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "CO2 (ppm)")

# only school hours 
dataset_final_regression_M_probandnorm %>%
   filter(name=="b_voc") %>%
     filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_boxplot(aes(y=mean_indoors, x=period_new,fill=class_type)) + 
   theme_minimal() +
    labs(y= "VOCs (ppb)")



```

## Period means for all schools (incl. problematic classrooms)

Periods: 

+ 1-10 = school periods from classes start time (7:45) to classes end time (13:05) e.g. 1 = 07:45 - 08:25
+ bef_classes = period from 7:00-7:45 
+ after_classes = period from 13:05-15:00 


+ 95th -ile

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

 # only school hours 
dataset_final_regression %>%
   filter(name=="pm10_0") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "PM10 (ug/m3)", x = "",
         title = "Mean PM10 values per school period",
         caption = "Purple colour: school break periods")

dataset_final_regression %>%
   filter(name=="pm1_0") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "PM1 (ug/m3)", x = "",
         title = "Mean PM1 values per school period",
         caption = "Purple colour: school break periods")

dataset_final_regression %>%
   filter(name=="pm2_5") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
      geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "PM2.5 (ug/m3)", x = "",
         title = "Mean PM2.5 values per school period",
         caption = "Purple colour: school break periods")


dataset_final_regression_M %>%
   filter(name=="temperature") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
      geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "Temperature (C)", x = "",
         title = "Mean Temperature values per school period",
         caption = "Purple colour: school break periods")
 
dataset_final_regression_M %>%
   filter(name=="humidity") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
      geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "Rel. Humidity (%)", x = "",
         title = "Mean Humidity values per school period",
         caption = "Purple colour: school break periods")

dataset_final_regression_M %>%
   filter(name=="carbon_dioxide") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
      geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "C02 (ppm)", x = "",
         title = "Mean CO2 values per school period",
         caption = "Purple colour: school break periods")

dataset_final_regression_M %>%
   filter(name=="b_voc") %>%
   filter(time_type != "bef_classes") %>% 
   filter(time_type != "after_classes") %>% 
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
      geom_point(aes(y=mean_indoors, x=period_start_time,  colour="indoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_point(aes(y=mean_outdoors, x=period_start_time,  colour="outdoors", shape = time_type),
              position=position_jitter(h=0.1, w=0.1),
             shape = 21, alpha = 0.5, size = 2) +
   geom_smooth(aes(x=period_start_time, y=mean_indoors, colour="indoors"), method="loess", span=0.1, se = F)+
      geom_smooth(aes(x=period_start_time, y=mean_outdoors, colour="outdoors"), method="loess", span=0.1, se = F)+
     #scale_x_datetime(labels = function(x) format(x, format = "%H:%M")) +
     #scale_x_datetime(date_label = "%H:%M") +
     #scale_x_datetime(date_breaks = "1 hour", labels = date_format("%H:%M"),
   #              timezone = "America/Toronto") +
   theme_minimal() +
   #annotate break times 09:05-09:25, 10:45-10:55, 12:15-12:25
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 09:05:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
   geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 10:45:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
      geom_vline(xintercept = as.numeric(as.POSIXct("2020-01-01 12:15:00")), color=RColorBrewer::brewer.pal(n = 6, "Set2")[4], size = 4, alpha=0.1) +
    labs(y= "VOCs (ppb)", x = "",
         title = "Mean VOC values per school period",
         caption = "Purple colour: school break periods")
```

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

# only school hours 

dataset_final_regression %>%
   filter(name=="pm10_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "PM10 (ug/m3)", x = "period")

dataset_final_regression %>%
   filter(name=="pm1_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "PM1 (ug/m3)", x = "period")

 # only school hours 
dataset_final_regression %>%
   filter(name=="pm2_5") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "PM2.5 (ug/m3)")

 # only school hours 
dataset_final_regression_M %>%
   filter(name=="temperature") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
       theme_minimal() +
labs(y= "Temperature (C)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="humidity") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "Rel. Humidity (%)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="carbon_dioxide") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "CO2 (ppm)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="b_voc") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour=location)) +
      theme_minimal() +
 labs(y= "VOCs (ppb)", x = "period")

 # only school hours 
dataset_final_regression %>%
   filter(name=="pm10_0") %>%
   filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "PM10 (ug/m3) 95th %ile", x = "period")

 # only school hours 
dataset_final_regression %>%
   filter(name=="pm1_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "PM1 (ug/m3)", x = "period")

 # only school hours 
dataset_final_regression %>%
   filter(name=="pm2_5") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "PM2.5 (ug/m3)", x = "period")

 # only school hours 
dataset_final_regression_M %>%
   filter(name=="temperature") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "Temperature (C)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="humidity") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "Rel. Humidity (%)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="carbon_dioxide") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
  geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "CO2 (ppm)", x = "period")

# only school hours 
dataset_final_regression_M %>%
   filter(name=="b_voc") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=period_new,  colour="mean_indoors")) +
   geom_point(aes(y=mean_outdoors, x=period_new,  colour="mean_outdoors")) +
      theme_minimal() +
    labs(y= "VOCs (ppb)", x = "period")

```

## Hourly means for all classrooms (incl. problematic classrooms) - all data

Periods: 

+ 1-24 = hours of day starting from 00:00 until 23:59 e.g. 1 = 00:00 - 00:59


+ 95th -ile

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

dataset_final_regression_all <- dataset_final_regression_all %>% 
   mutate(hour=as.numeric(hour))


dataset_final_regression_all %>%
   filter(name=="pm10_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
      theme_minimal() +
 labs(y= "PM10 (ug/m3)")

 # only school hours 
dataset_final_regression_all %>%
   filter(name=="pm1_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
     theme_minimal() +
  labs(y= "PM1 (ug/m3)")

 # only school hours 
dataset_final_regression_all %>%
   filter(name=="pm2_5") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
       theme_minimal() +
labs(y= "PM2.5 (ug/m3)")

 # only school hours 
dataset_final_regression_all_M %>%
   filter(name=="temperature") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour, color=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
      theme_minimal() +
 labs(y= "Temperature (C)")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="humidity") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
      theme_minimal() +
 labs(y= "Rel. Humidity")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="carbon_dioxide") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
     theme_minimal() +
  labs(y= "CO2 (ppm)")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="b_voc") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour=location)) +
      geom_smooth(aes(x=hour, y=mean_indoors, colour=location), size = 1, method="loess", span=0.1,se = F) +
       theme_minimal() +
labs(y= "VOCs (ppb)")



dataset_final_regression_all %>%
   filter(name=="pm10_0") %>%
 filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
      theme_minimal() +
 labs(y= "PM10 (ug/m3)")

 # only school hours 
dataset_final_regression_all %>%
   filter(name=="pm1_0") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
      theme_minimal() +
 labs(y= "PM1 (ug/m3)")

 # only school hours 
dataset_final_regression_all %>%
   filter(name=="pm2_5") %>%
    filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
      theme_minimal() +
 labs(y= "PM2.5 (ug/m3)")

 # only school hours 
dataset_final_regression_all_M %>%
   filter(name=="temperature") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
      theme_minimal() +
 labs(y= "Temperature (C)")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="humidity") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
   theme_minimal() +
    labs(y= "Rel. Humidity")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="carbon_dioxide") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
      theme_minimal() +
 labs(y= "CO2 (ppm)")

# only school hours 
dataset_final_regression_all_M %>%
   filter(name=="b_voc") %>%
      filter(mean_indoors<quantile(mean_indoors, probs=0.95, na.rm=TRUE)) %>%
   filter(mean_outdoors<quantile(mean_outdoors, probs=0.95, na.rm=TRUE)) %>%
   ggplot() +
   geom_point(aes(y=mean_indoors, x=hour,  colour="indoors")) +
   geom_smooth(aes(x=hour, y=mean_indoors, colour="indoors"), size = 1, method="loess", span=0.1,se = F)+
   geom_point(aes(y=mean_outdoors, x=hour,  colour="outdoors")) +
   geom_smooth(aes(x=hour, y=mean_outdoors, colour="outdoors"), size = 1, method="loess", span=0.1,se = F)+
   theme_minimal() +
    labs(y= "VOCs (ppb)")
```

# Session Information

```{r S7 }
sessionInfo()
```