Plotting FAIR metrics

FAIR_metrics.Rmd

---
title: "FAIR Metrics"
author: "Matt Jones"
date: "5/8/2019"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(dplyr)
library(tidyr)
library(ggplot2)
library(scales)
```

## Generate a simulated data set

This is a fake data set that takes the form:

```{r load_data}
updates <- data.frame(v1=seq(as.Date("2000/1/1"), by = "month", length.out = 10),
                      v2=seq(as.Date("2000/5/1"), by = "month", length.out = 10),
                      v3=seq(as.Date("2000/8/1"), by = "month", length.out = 10)) %>% 
    gather(key = version, value = update_date) %>% 
    arrange(update_date) %>% 
    select(update_date)
fair <- expand.grid(version = seq(1,3,1), object = seq(1,5,1), scope = c("adc", "knb")) %>% 
    mutate(pid = paste(scope, object, version, sep=".")) %>%
    mutate(f = sample(90:100, 30, replace=TRUE)) %>%
    mutate(a = sample(70:100, 30, replace=TRUE)) %>%
    mutate(i = sample(40:70, 30, replace=TRUE)) %>%
    mutate(r = sample(30:50, 30, replace=TRUE)) %>%
    mutate(score = as.integer((f + a + i + r)/4))
scores <- bind_cols(updates, fair)
head(scores)
```

## Calculate stats

Generate stats by first grouping by month, then keep only the most recent 
observation for each dataset that month, and then take the average of those 
for each of the FAIR facets by month.  Finally, transpose the data.
```{r calculate_means}
most_recent <- scores %>%
    arrange(update_date, object, version) %>%
    group_by(update_date, object) %>% 
    top_n(1, version)

score_means <- most_recent %>%
    group_by(update_date) %>% 
    summarise(f=mean(f), a=mean(a), i=mean(i), r=mean(r)) %>%
    gather(metric, mean, -update_date)
score_means$metric <- factor(score_means$metric, 
    levels=c("f", "a", "i", "r"), 
    labels=c("Findable", "Accessible", "Interoperable", "Reusable"))
head(score_means)
```

## Plot it!

```{r plot}
d1_colors <- c("#ff582d", "#c70a61", "#1a6379", "#60c5e4")
ggplot(data=score_means, mapping=aes(x=update_date, y=mean, color=metric)) +
    geom_line() +
    geom_point(size=1) +
    theme_bw() +
    scale_colour_manual(values=d1_colors) +
    scale_x_date(date_breaks="3 months", date_minor_breaks="months", labels=date_format("%Y %b")) +
    scale_y_continuous(limits=c(0,100))
```