dv-pred.qmd

---
fig-height: 4.5
fig-width: 4.5
---

```{r setup, include = FALSE}
source("global.R")
```

# Observed vs predicted

We define a variable that can be passed as `yname`, or the name of `DV`

```{r}
.yname
```

## DV versus PRED `[dv_pred]`

```{r}
dv_pred(df, yname = .yname)
```


### Un-fix x- and y-axis limits

By default, pmplots forces the x- and y- scales to have the same limits so you 
get a square plot. You can prevent this behavior with 

```{r}
dv_pred(df, scales = "free")
```

If you are faceting, you'll have to pass `scales="free"` to both `dv_pred()` and 
`facet_wrap()`

```{r}
dv_pred(df, scales = "free") + facet_wrap(~RF, scales = "free")
```



## DV versus PRED - log/log {.unnumbered .unlisted}

```{r}
dv_pred(df, loglog=TRUE, yname = .yname)
```

## DV versus IPRED `[dv_ipred]`

```{r}
dv_ipred(df, yname = .yname)
```

## DV versus IPRED - log/log {.unnumbered .unlisted}

```{r}
dv_ipred(df, loglog = TRUE, yname = .yname)
```

## DV versus both PRED and IPRED `[dv_preds]`

```{r}
#| fig-width: 8
dv_preds(df, yname = .yname) %>% pm_grid(ncol = 2)
```