---
title: "7. Answers to common questions and extra examples"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{7. Misc. questions}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
library(cubble)
library(dplyr)
```

This vignette documents miscellaneous examples and questions received regarding 
the `cubble` package.

# Why summarising of temporal data by station is not in temporal form?

Q: I'm trying to summarise the average maximum temperature by station. I'm not sure why the summarise function is not returning a temporal object. 
  
```{r}
climate_mel |> 
  face_temporal() |> 
  summarise(tmax_avg = mean(tmax, na.rm=TRUE))
```

**A: This operation should be performed in the spatial form. The form to use for an operation depends on the structure of the result. If the result has each key in a row without temporal index, it should be operated in the spatial form (example here):**

```{r}
climate_mel |> 
  rowwise() |> 
  mutate(tmax_avg = mean(ts$tmax, na.rm=TRUE))
```

The same treatment applies, if your spatial cubble is also an sf object.