120-iteration.Rmd

# Iteration

## Broadcasting

```{r}
f_values <- c(0, 32, 212, -40)
```

```{r}
f_values * 10
```

```{r}
f_values * c(10, 100)
```

## For loops

Temp conversion functions

```{r}
f_k <- function(f_temp) {
    ((f_temp - 32) * (5 / 9)) + 273.15
}


k_c <- function(temp_k) {
    if (is.na(temp_k)) {
        return(NA)
    } else if (temp_k < 0) {
        warning('you passed in a negative Kelvin number')
        # stop()
        return(NA)
    } else {
        temp_c <- temp_k - 273.15
        return(temp_c)
    }
}

f_c <- function(temp_f) {
    temp_k <- f_k(temp_f)
    temp_c <- k_c(temp_k)
    return(temp_c)
}
```



```{r}
for (pizza in f_values) {
    print(pizza)
    converted <- f_c(pizza)
    print(converted)
}
```

```{r}
# 1:length(f_values)
# seq_along(f_values)
for (i in seq_along(f_values)) {
    print(i)
    val <- f_values[i]
    print(val)
    
    converted <- f_c(val)
    print(converted)
}
```

### Pre allocating in a loop

```{r}
# prepopulate an empty vector
converted_values <- vector("double", length(f_values))
for (to_be_converted_position in seq_along(f_values)) {
    converted <- f_c(to_be_converted_position)
    converted_values[to_be_converted_position] <- converted
}
```


```{r}
converted_values
```

## purrr (map)

```{r}
library(purrr)
```

```{r}
map(f_values, f_c)
```

```{r}
map_dbl(f_values, f_c)
```

```{r}
mtcars
```

```{r}
map(mtcars, class)
```

```{r}
map_chr(mtcars, class)
```


```{r}
map_dbl(mtcars, mean)
```

```{r}
map(mtcars, summary)
```

## Apply (in base R)

apply family of functions


### lapply

```{r}
lapply(f_values, f_c)
```

### sapply

```{r}
sapply(f_values, f_c)
```

```{r}
v1 <- c(1, 2, 3, 4)
v2 <- c(100, 200, 300, 400)
```

### mapply

```{r}
my_mean <- function(x, y){
    return((x + y) / 2)
}
```

```{r}
# sapply(v1, v2, my_mean)
```

```{r}
mapply(my_mean, v1, v2)
# this is the same as purrr::map2
```

### apply (2-dimensions)

```{r}
apply(mtcars, MARGIN = 1, mean)
```

```{r}
apply(mtcars, MARGIN = 2, mean)
```