Add some sample code analysing data from the app (#119)

* working R test script plotting MoE's

* add readme

* remove stuff specific to MoE

* remove margins of error

* clean + comment code

* refresh data from live app

analysis/MoE-test.r

@@ -0,0 +1,75 @@
+library('tidyverse')
+
+setwd('C:\\Users\\nwessel\\Documents\\GitHub\\tt_app\\analysis')
+
+purple = '#9e7cbd'
+grey = '#4f4c4c'
+
+# compare monthly data for two periods by direction of travel
+read_csv('results-monthly.csv') %>%
+    group_by( dateRange, direction) %>%
+    mutate(ttt = mean_travel_time_minutes) %>%
+    extract(
+        dateRange,
+        into=c('fromDate'),
+        regex='From ([0-9]{4}-[0-9]{2}-[0-9]{2})'
+    ) %>%
+    mutate( fromDate= ymd(fromDate) + days(15) ) %>%
+    ggplot( aes( x=fromDate, y=ttt, lty=direction ) ) +
+        geom_line( size = 1.2, color=grey ) +
+        geom_point( color=grey, size=2 ) +
+        annotate( # approximate start/end dates of complete-street installation
+            'rect',
+            xmin = ymd('2023-09-11'),
+            xmax = ymd('2023-10-15'),
+            ymin = -Inf,
+            ymax = Inf,
+            fill = grey,
+            alpha = 0.2
+        ) +
+        scale_y_continuous(
+            breaks = c(0,2,4,6,8,10,12),
+            limits = c(0, NA)
+        ) +
+        scale_x_date(
+            breaks = '1 month',
+            minor_breaks = '1 month',
+            labels = scales::label_date_short()
+        ) +
+        labs( title = 'Bloor West weekday travel times between Runnymede & Aberfoyle (PM Peak)' ) + 
+        ylab( 'Monthly Average Travel Time (minutes)' ) +
+        xlab( 'Month' ) +
+        theme_bw()
+
+
+# compare two date ranges plotted by hour of day
+read_csv('results-ToD.csv') %>%
+    extract(
+        timeRange,
+        into=c('hour'),
+        regex='From ([0-9]+):00'
+    ) %>%
+    mutate(
+        hour = as.numeric(hour),
+        period = factor(case_match(
+            dateRange,
+            'From 2022-11-01 to 2023-04-01' ~ 'before',
+            'From 2023-11-01 to 2024-04-01' ~ 'after'
+        ), levels=c('before','after')),
+    ) %>%
+    ggplot( aes( x=0.5+hour, y=mean_travel_time_seconds, lty=direction ) ) +
+    geom_line( aes(color=period,), size = 1 ) +
+    scale_color_manual(values=c('before'=grey,'after'=purple)) +
+    labs( title = 'Bloor West travel times between Runnymede & Aberfoyle' ) + 
+    ylab( 'Travel Time (seconds)' ) +
+    xlab( 'Hour of day' ) + 
+    scale_y_continuous(
+        breaks = 60 * seq(0,14,1),
+        minor_breaks = NULL,
+        limits = c(0, NA)
+    ) +
+    scale_x_continuous(
+        breaks = seq(0,24),
+        minor_breaks = c(),
+    ) +
+    theme_bw()

analysis/readme.md

@@ -0,0 +1,3 @@
+# Analysing data from the app
+
+This folder contains some scripts that demo how to work with data from the travel time app.