Testing_running_time_in_R.qmd

# Testing R performance versus SAS

## Task 1
What are the counts and proportions of males and females in outpatient claims (2017-2019) by year?

```{r}
tic()
ccaeo |> 
  filter(YEAR %in% 2017:2019) |> 
  select(SEX, YEAR) |> 
  group_by(YEAR) |> 
  count(SEX) |> 
  to_duckdb()
toc()
```

## Task 2 

Calculate summary statistics for the co-payments in pharmacy claims (2017-2019) for metformin-containing medications
```{r}
tic()
ccaed |> 
  select(ENROLID, NDCNUM, COPAY, YEAR) |> 
  filter(YEAR %in% 2017:2019) |> 
  filter(!is.na(ENROLID)) |> 
  select(-ENROLID) |> 
  filter(str_detect(NDCNUM, metformin_drugs_regex)) |> 
  select(-NDCNUM) |> 
  to_duckdb() |> 
  collect() |> 
  summarise(COPAY2 = mean(COPAY, na.rm = T), .by = YEAR)
toc()
```


## Task 3
How many pioglitazone dispensing claims with at least a 30-day supply in 2018 and 2019?
```{r}
tic()
ccaed |> 
  select(ENROLID, NDCNUM, YEAR, DAYSUPP) |> 
  filter(YEAR %in% 2018:2019) |> 
  filter(!is.na(ENROLID), DAYSUPP >= 30) |> 
  select(-ENROLID) |> 
  filter(str_detect(NDCNUM, pioglitazone_drugs_regex)) |> 
  select(-NDCNUM) |> 
  to_duckdb() |> 
  group_by(YEAR) |> 
  count()
toc()
```

## Task 4
Calculate the summary statistics of the number of hospital admissions per beneficiary (2017-2019).

```{r}
tic()
ccaei |> 
  select(ENROLID, ADMDATE, YEAR) |> 
  filter(YEAR %in% 2017:2019) |> 
  filter(!is.na(ENROLID)) |> 
  distinct() |> 
  count(ENROLID, sort = T) |> 
  collect()
toc()
```



## Task 5
Calculate the counts and proportions of type 2 diabetes patients by region (2017-2019)

```{r}
tic()
ccaeo |> 
  filter(!is.na(ENROLID), YEAR %in% 2017:2019) |> 
  select(ENROLID, contains("DX"), REGION) |> 
  to_duckdb() |> 
  filter(
    (if_any(contains("DX"), ~str_detect(.x, {{type2_diabetes_regex}})))
  ) |> 
  distinct(ENROLID, REGION) |> 
  count(REGION)
toc()
```