index.Rmd

---
title: "NSF HNDS Awards"
author: "Rick Gilmore"
date: "`r Sys.Date()`"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(tidyverse)
```

## Purpose

This document provides details about awards made under the U.S. National Science Foundation (NSF) [Human Networks and Data Science](https://beta.nsf.gov/funding/opportunities/human-networks-data-science-hnds) program.

## Acquire award data

Awards made under the program can be found [here](https://www.nsf.gov/awardsearch/advancedSearchResult?ProgEleCode=130Y%2C147Y&BooleanElement=Any&BooleanRef=Any&ActiveAwards=true#results).

The API *should* permit a CSV formatted download of the awards data, but none of the following worked for me.
I get a '500 Internal Server Error'.

```{r, eval=FALSE}
download.file("https://www.nsf.gov/awardsearch/ExportResultServlet?exportType=csv", destfile = "csv/hnds.csv")
hdns <- read.csv("https://www.nsf.gov/awardsearch/ExportResultServlet?exportType=csv")
hnds <- readr::read_csv("https://www.nsf.gov/awardsearch/ExportResultServlet?exportType=csv")
```

Rather than perseverate on making the download automatic, I'm just copying the downloaded file to the local directory and moving on.

## Load

```{r}
hnds <- readr::read_csv("csv/Awards.csv", show_col_types = FALSE)
```

## Explore

Examine fields.

```{r}
names(hnds)
```

## Clean

```{r}
hnds <- hnds %>%
  dplyr::mutate(., AwardedToDate = stringr::str_remove(AwardedAmountToDate, "\\$|,")) %>%
  dplyr::mutate(., AwardedToDate = stringr::str_remove_all(AwardedToDate, ",")) %>%
  dplyr::mutate(., AwardedToDate = as.numeric(AwardedToDate))
```

## Histogram of award size

```{r}
hnds %>%
  ggplot() +
  aes(x = AwardedToDate) +
  geom_histogram()
```

## Focus on infrastructure grants

```{r, results='hide'}
hnds$`Program(s)`
```

If the `Program(s)` field contains 'Human Networks & Data Sci Infr', the award is an infrastructure award.

```{r}
hnds <- hnds %>%
  dplyr::mutate(., hnds_inf = stringr::str_detect(`Program(s)`, "Human Networks & Data Sci Infr"))
```

```{r}
hnds %>%
  dplyr::filter(., hnds_inf) %>%
  dplyr::arrange(., desc(AwardedToDate)) %>%
  dplyr::select(., Title, PrincipalInvestigator, Organization, AwardedAmountToDate) %>%
  knitr::kable(format = 'html') %>%
  kableExtra::kable_classic()
```

These are total award amounts, so to approximate the size of the direct costs, we divide by 1.6.

```{r}
hnds <- hnds %>%
  dplyr::mutate(., est_direct_cost = AwardedToDate/1.6)
```

```{r}
hnds %>%
  dplyr::filter(., hnds_inf) %>%
  dplyr::arrange(., desc(AwardedToDate)) %>%
  dplyr::select(., Title, PrincipalInvestigator, Organization, AwardedAmountToDate, est_direct_cost) %>%
  knitr::kable(format = 'html') %>%
  kableExtra::kable_classic()
```