cop-dashboard.Rmd

---
title: "Data Science Community of Practice"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    vertical_layout: scroll
    source_code: https://github.com/bcgov/bcgov-data-science-cop
---

<!--
Copyright 2020 Province of British Columbia

This work is licensed under the Creative Commons Attribution 4.0 International License.
To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
-->


```{r setup, include=FALSE}
## libraries
library(flexdashboard)
library(readr)
library(janitor)
library(lubridate)
library(dplyr)
library(tidyr)
library(ggplot2)
library(waffle) #dev version remotes::install_github("hrbrmstr/waffle")
library(RColorBrewer)
library(plotly)
library(safepaths) ## https://github.com/bcgov/safepaths
library(reactable)
library(htmltools)
library(crosstalk)
library(flextable)
library(treemap)


## load data
event_participation <-
  read_csv(use_network_path("7. Data Science CoP/data/cop-data-events.csv"),
           col_types = c("ccddDccdc")) %>%
  clean_names()

participation_by_ministry <-
  read_csv(use_network_path("7. Data Science CoP/data/cop-data-part-ministries.csv"),
           col_types = c("cDdcc")) %>%
  clean_names()

workshop_feedback <-  read_csv(use_network_path("7. Data Science CoP/data/cop-training-survey-feedback.csv"),
           col_types = c("ccdDccddddddddddddcc")) %>%
  clean_names()

org_acronyms <- read_csv(use_network_path("7. Data Science CoP/data/ministry-name-abbrevation.csv"),
           col_types = c("cc")) %>%
  clean_names()
```


Event Participation
======================================================================


Row 
-----------------------------------------------------------------------


### Number of Events (All Time) {.value-box}

```{r}
#number of cop events
valueBox(value = event_participation %>% 
  count(), icon = "fa-calendar")
```

### Total Attendance (All Time) {.value-box}

```{r}
#cumulative event attendance
participation_count <- event_participation %>%
  filter(date < today()) %>%
  select(date, event_title, event_type, in_person_participants, on_line_participants) %>%
  mutate(on_line_participants = replace_na(on_line_participants, 0),
         in_person_participants = replace_na(in_person_participants, 0),
         participants = in_person_participants + on_line_participants)

valueBox(value = participation_count %>% 
  summarise(sum(participants)), icon = "fa-users")
```

### Geographic Locations {.value-box}

```{r}
#number of locations
num_locations <- event_participation %>% 
  distinct(location) %>% 
  filter(location != "remote") %>% 
  count()
 
valueBox(value = num_locations, icon = "fa-globe")
```

### Number of Ministries & Organizations Represented (All Time) {.value-box}

```{r}
#number of ministries
valueBox(value = participation_by_ministry %>% 
          filter(!ministry %in% c("Unknown", "External")) %>% 
  distinct(ministry) %>% 
  count(), icon = "fa-building")
```


Row 
-----------------------------------------------------------------------


### Number of Events by Year

```{r}
#number of events by year and event type
event_sum <- participation_count %>%
  mutate(year = year(date)) %>% 
  mutate(event_type = recode(event_type, social = "meet-up")) %>% 
  group_by(year, event_type) %>% 
  count()

#event colours
event_colours = c("meet-up" = "#1f78b4",
                  "webinar" = "#ff7f00",
                  "workshop" = "#33a02c")

#event bar chart theme
theme_bar <- theme(
    panel.grid.major.x = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.line = element_line(),
    axis.text = element_text(size = 11),
    legend.position = "top",
    legend.direction = "horizontal",
    legend.text = element_text(size = 11))

#bar chart of number of events
ggplot(data = event_sum, aes(year, n)) +
  geom_col(aes(fill = event_type), alpha = 0.6) +
  labs(
    x = NULL,
    y = NULL
  ) +
  scale_y_continuous(expand = c(0, 0),
                     limits = c(0, 22),
                     breaks = seq(0, 22, 4)) +
  scale_fill_manual(name = NULL,
                    values = event_colours) +
  theme_minimal() +
  theme_bar
```

### Locations of In-Person Events

```{r}
#get cities from bcmaps
cities <- bcmaps::bc_cities(class = "sf") %>% 
  filter(NAME %in% c("Vancouver", "Victoria", "Prince George"))

#join cities to event data
event_locations <- event_participation %>% 
  select(location) %>% 
  group_by(location) %>% 
  count() %>% 
  ungroup() %>% 
  left_join(cities, by = c("location" = "NAME")) %>% 
  # mutate(location = recode(location, "Prince George" = "Prince\nGeorge")) %>% 
  sf::st_as_sf()

#plot in-person event locations
event_locations %>% 
  filter(location != "remote") %>% 
  ggplot() +
  geom_sf(data = bcmaps::bc_bound(class = "sf"), fill = "white") +
  geom_sf(colour = "#33a02c", size = 3) +
  geom_sf_text(aes(label = location),
               vjust = 1.8, hjust = -.01,
               size = 3.5, colour = "grey40") +
  # coord_sf(datum = NA) +
  labs(x = NULL, y = NULL) +
  theme_minimal() +
  theme(legend.position = "none")
```


Row
-----------------------------------------------------------------------


### Event Attendance by Year

```{r}
#participation by year and event
participation_sum <- participation_count %>%
  mutate(year = year(date)) %>% 
  mutate(event_type = recode(event_type, social = "meet-up")) %>% 
  group_by(year, event_type) %>% 
  summarise(total_part = sum(participants),
            total_event = length(event_type)) 

#bar chart of number of participants
ggplot(data = participation_sum, aes(year, total_part)) +
  geom_col(aes(fill = event_type), alpha = 0.6) +
  labs(
    x = NULL,
    y = NULL
  ) +
  scale_y_continuous(expand = c(0, 0),
                     limits = c(0, 600),
                     breaks = seq(0, 600, 100)) +
  scale_fill_manual(name = NULL,
                    values = event_colours) +
  theme_minimal() +
  theme_bar
```

### Attendance by Ministry or Organization

```{r}
#waffle colour palette
# colour_count_waffle <-  length(unique(participation_by_ministry$ministry))
# get_palette_waffle <-  colorRampPalette(brewer.pal(12, "Paired"))

#waffle plot
# participation_by_ministry %>%
#   left_join(org_acronyms, by = "ministry") %>% 
#   filter(!ministry %in% c("Unknown", "External")) %>% 
#   # mutate(ministry = stringr::str_wrap(ministry, 45)) %>% 
#   group_by(abbreviation) %>%
#   summarise(count = sum(number_participants)) %>%
#   ggplot(aes(fill = abbreviation, values = count)) +
#   geom_waffle(rows = 10) +
#   scale_fill_manual(values = get_palette_waffle(colour_count_waffle),
#                     name = NULL) +
#   coord_equal() +
#   theme_enhance_waffle() +
#   theme_void() +
#   theme(
#     legend.position = "bottom",
#     legend.text = element_text(size = 8)) +
#   guides(fill = guide_legend(nrow = 6))

#treemap
df <- participation_by_ministry %>%
  left_join(org_acronyms, by = "ministry") %>%
  filter(!ministry %in% c("Unknown", "External")) %>%
  group_by(abbreviation) %>%
  summarise(count = sum(number_participants))

treemap(
  df,
  index = "abbreviation",
  vSize = "count",
  type = "index",
  fontsize.labels = 8,
  force.print.labels = TRUE,
  fontface.labels = "plain",
  border.lwds = .5,
  title = "",
  palette = brewer.pal(12, "Paired")
)
```


Training Events
======================================================================


Row
-----------------------------------------------------------------------


### Number of Training Events (Total All Time) {.value-box}

```{r}
# number of workshops + webinars delivered all time
valueBox(value = event_participation %>% 
  filter(event_type %in% c("workshop", "webinar")) %>% 
  count(), icon = "fa-calendar")
```

### Number of Training Attendees (Total All Time) {.value-box}

```{r}
#number of workshop+webinars attendees all time
valueBox(value = participation_count %>% 
  filter(event_type %in% c("workshop", "webinar")) %>% 
  summarise(sum(participants)), icon = "fa-users")
```


Row
-----------------------------------------------------------------------


### Workshops (Total All Time) {.value-box}

```{r}
# number of workshops delivered all time
valueBox(value = event_participation %>% 
  filter(event_type == "workshop") %>% 
  count(), icon = "fa-building")
```

### Webinars (Total All Time) {.value-box}

```{r}
# number of webinars delivered all time
valueBox(value = event_participation %>% 
  filter(event_type == "webinar") %>% 
  count(), icon = "fa-video")
```

### Workshop Training Attendees (Total All Time) {.value-box}

```{r}
#number of workshop attendees all time
valueBox(value = participation_count %>% 
  filter(event_type == "workshop") %>% 
  summarise(sum(participants)), icon = "fa-user")
```

### Webinar Training Attendees (Total All Time) {.value-box}

```{r}
#number of webinar attendees all time
valueBox(value = participation_count %>% 
  filter(event_type == "webinar") %>% 
  summarise(sum(participants)), icon = "fa-user")
```


Row
-----------------------------------------------------------------------


### Number of Training Events by Year

```{r}
#number of events by year and event type
training_sum <- participation_count %>%
  mutate(year = year(date)) %>% 
  filter(event_type %in% c("workshop", "webinar")) %>% 
  group_by(year, event_type) %>% 
  count()

#training event colours
training_colours = c("webinar" = "#ff7f00",
                  "workshop" = "#33a02c")

#bar chart of number of events
ggplot(data = training_sum, aes(year, n)) +
  geom_col(aes(fill = event_type), alpha = 0.6) +
  labs(
    x = NULL,
    y = NULL
  ) +
  scale_y_continuous(expand = c(0, 0),
                     limits = c(0, 12),
                     breaks = seq(0, 12, 2)) +
  scale_fill_manual(name = NULL,
                    values = training_colours) +
  theme_minimal() +
  theme_bar
```

### Training Attendance by Event (Number Offerings)

```{r}
#event horizontal bar chart theme
theme_bar_horiz <- theme(
    panel.grid.major.y = element_blank(),
    panel.grid.minor.y = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.line = element_line(),
    axis.text = element_text(size = 11),
    legend.position = "top",
    legend.direction = "horizontal",
    legend.text = element_text(size = 11))

#number participants by topic
participation_count %>%
  filter(event_type %in% c("workshop", "webinar")) %>% 
  mutate(event_title2 = case_when(event_type == "webinar" ~ "data-science-cop-webinar",
                                  TRUE ~ event_title)) %>% 
  group_by(event_title2, event_type) %>% 
  summarise(total_trained = sum(participants),
            no_times_offered = n()) %>% 
  ggplot(aes(total_trained, reorder(event_title2, -total_trained))) +
  geom_col(aes(fill = event_type), alpha = 0.6) +
  scale_fill_manual(name = NULL,
                    values = training_colours) +
  geom_text(aes(label = paste0("(", no_times_offered, ")")),
            nudge_x = 50, colour = "grey30", size = 4) +
  labs(x = NULL, y = NULL) +
  scale_x_continuous(expand = c(0, 0),
                     limits = c(0, 950),
                     breaks = seq(0, 900, 150)) +
  theme_minimal() +
  theme_bar_horiz
```


Row
-----------------------------------------------------------------------


### Workshop Evaluation: Rating of Excellent + Very Good (Avg All Workshops)

```{r}
#percent excellent + very good overall assessment scores (averaged across workshops)
metric_overall <- workshop_feedback %>% 
  mutate(metric = overall_assessment_excellent_percent + overall_assessment_very_good_percent) %>% 
  select(metric) %>% 
  summarise(percent = mean(metric, na.rm =TRUE)) %>% 
  mutate(percent = paste0(percent, "%")) %>%
  pull()

gauge(metric_overall, min = 0, max = 100, symbol = '%', gaugeSectors(
  success = c(80, 100), warning = c(40, 79), danger = c(0, 39)
))

# valueBox(value = metric_overall, icon = "fa-comments",
#          caption = "Excellent or Very Good Overall Rating (All Workshops)")
```

### Workshop Organization Evaluation: Rating of Excellent + Very Good (Avg All Workshops)

```{r}
#percent excellent + very good organisation assessment scores (averaged across workshops)
metric_organization <- workshop_feedback %>% 
  mutate(metric = organization_assessment_excellent_percent + organization_assessment_very_good_percent) %>% 
  select(metric) %>% 
  summarise(percent = mean(metric, na.rm =TRUE)) %>% 
  mutate(percent = paste0(percent, "%")) %>% 
  pull()

gauge(metric_organization, min = 0, max = 100, symbol = '%', gaugeSectors(
  success = c(80, 100), warning = c(40, 79), danger = c(0, 39)
))

# valueBox(value = metric_organization, icon = "fa-commenting-o",
#          caption = "Excellent or Very Good Organization Rating (All Workshops)")
```


Row
-----------------------------------------------------------------------


### Workshop Participant Assessments: Overall Rating by Workshop

```{r}
tidy_feedback <- workshop_feedback %>%
  pivot_longer(cols = starts_with(c("overall", "organization")),
               names_to = "category",
               values_to = "percent") %>%
  select(event_title, category, percent) %>% 
  mutate(cat_label = stringr::str_remove_all(category, "(overall|organization)_assessment_"),
         cat_label = stringr::str_remove(cat_label, "_percent"),
         cat_label = stringr::str_replace(cat_label, "_", " "),
         cat_label = factor(cat_label, levels = c("excellent",
                                                  "very good",
                                                  "good",
                                                  "fair",
                                                  "poor")))

#colour palette
colour_count <-  length(unique(tidy_feedback$cat_label))
get_palette <-  colorRampPalette(brewer.pal(9, "YlGn")[2:9])

# colour_palette <- c("poor" = "pink",
#                     "fair" = "red",
#                     "good" = "green",
#                     "very good" = "blue",
#                     "excellent" = "black")

p <- tidy_feedback %>% 
  drop_na() %>% 
  filter(stringr::str_starts(category, "overall")) %>% 
  group_by(event_title, cat_label) %>% 
  summarise(avg_percent = mean(percent)) %>% 
  ggplot(aes(avg_percent, event_title, fill = cat_label)) +
  geom_col(alpha = 0.6) +
  scale_x_continuous(expand = c(0, 0),
                     limits = c(0, 110),
                     breaks = seq(0, 100, 20)) +
  scale_fill_manual(values = rev(get_palette(colour_count)),
                    name = NULL,
                    guide = guide_legend(reverse = TRUE),
                    na.value = "black") +
  labs(x = NULL, y = NULL) +
  theme_minimal() +
  theme_bar_horiz +
  theme(legend.position = "right",
        legend.direction = "vertical")
p

# ggplotly(p)
```

### Workshop Participant Assessments: Organization Rating by Workshop

```{r}
tidy_feedback %>% 
  drop_na() %>% 
  filter(stringr::str_starts(category, "organization")) %>% 
  group_by(event_title, cat_label) %>% 
  summarise(avg_percent = mean(percent)) %>% 
  ggplot(aes(avg_percent, event_title, fill = cat_label)) +
  geom_col(alpha = 0.6) +
  scale_x_continuous(expand = c(0, 0),
                     limits = c(0, 110),
                     breaks = seq(0, 100, 20)) +
  scale_fill_manual(values = rev(get_palette(colour_count)),
                    name = NULL,
                    guide = guide_legend(reverse = TRUE),
                    na.value = "black") +
  labs(x = NULL, y = NULL) +
  theme_minimal() +
  theme_bar_horiz +
  theme(legend.position = "right",
        legend.direction = "vertical")
```


Explore Event Data 
======================================================================


Row {data-height="900"}
-----------------------------------------------------------------------

```{r}
event_num_min <- participation_by_ministry %>% 
  select(event_title, date, ministry) %>% 
  distinct() %>% 
  group_by(date, event_title) %>% 
  count()

table_df <- participation_count %>%
  left_join(event_num_min, by = c("event_title", "date")) %>% 
  select(event_type, event_title, date, n, participants) %>%
  arrange(desc(date)) 

data <- SharedData$new(table_df)

bscols(
  widths = c(4,8),
  list(
    filter_checkbox("type", "Event Type", data, ~event_type),
    filter_slider("part", "Number of Participants", data, ~participants, width = "100%")
  ),
  reactable(data, 
            columns = list(
  event_type = colDef(name = "Event Type" #,
                      # footer = "Total"
                      ),
  date = colDef(name = "Date", format = colFormat(date = TRUE, locales = "en-GB")),
  event_title = colDef(name = "Event"),
  participants = colDef(name = "Number of Participants",
                        align = "center" #,
                        # footer = function(values) sum(values)
                        ),
  n = colDef(name = "Number of Ministries/Organizations",
             align = "center",
             na = "NA")),
  # defaultColDef = colDef(footerStyle = list(fontWeight = "bold")),
# filterable = TRUE,
width = 700,
height = "100%",
# pagination = FALSE,
showPageSizeOptions = TRUE,
# pageSizeOptions = c(10, 20, 50),
defaultPageSize = 10,
minRows = 10,
highlight = TRUE,
striped = TRUE)
)
```


```{r, include=FALSE, eval=FALSE}
library(officer)

border_format <- fp_border(color = "gray30", width = 2)

ft_table <- table_df %>%
  flextable() %>%
  set_header_labels(
    event_type = "Event Type",
    date =  "Date",
    event_title = "Event",
    participants = "Number of Participants",
    n = "Number Ministries Represented"
  ) %>%
  set_table_properties(width = .7, layout = "autofit") %>%
  theme_zebra(odd_header = "transparent", even_header = "transparent") %>%
  align(align = "left", part = "header") %>%
  align(align = "left", part = "body") %>%
  hline_bottom(part = "header", border = border_format) %>%
  hline_top(part = "header", border = border_format) %>%
  hline_bottom(part = "all", border = border_format)

ft_table
```


About
======================================================================


Row {data-height=75}
-----------------------------------------------------------------------

The **Data Science Community of Practice (CoP)** was launched in August 2018 as a venue for data science enthusiasts across the B.C. government to connect, communicate, share and learn about data science. The community exists online&mdash;and sometimes in-person&mdash;with a focus on sharing practices, inspiring each other and supporting continuous learning in data science in the BC Public Service.  

Row {data-height=200}
-----------------------------------------------------------------------

```{r, fig.width=3, fig.height=3}

knitr::include_graphics(use_network_path("7. Data Science CoP/CoP-logo-sticker/cop-logo-updated.png")) 
```

Row {data-height=100}
-----------------------------------------------------------------------

• **Join** the bcgov Data Science Community of Practice online in bcgov Yammer and MS Teams  
• **Read** blog posts about past & future bcgov Data Science Community of Practice events on @Work  
• **Find** bcgov Data Science Community of Practice materials & resources on [bcgov GitHub](https://github.com/bcgov/bcgov-data-science-cop)