- Uses Testthat Third edition, and implements snapshot testing

- Address bug with not creating equal strata
- Address warnings when using b_diff_summary, which now returns NA if there is only one observation, as we can’t take the difference of one observation, and a difference of 0 in these cases would be misleading.

- fix warning bug in
`keys_near`

related to factors - Add
`feat_diff_summary()`

functions to help summarise diff(). Useful for exploring the time gaps in the`index`

. (#100) - sample functions now work with multiple keys (#85, #89) (Thanks to @earowang and @deanmarchiori for their help with this.)
`facet_sample()`

now has a default of 3 per plot- resolve features(data ,.key, n_obs) error (#71)
- For
`near_quantile()`

, the`tol`

argument now defaults to 0.01. - provide an S3 generic for
`tbl_ts`

objects for`keys_near()`

- #76 - Add new dataset,
`pisa`

containing a short summary of the PISA dataset from https://github.com/ropenscilabs/learningtower for three (of 99) countries - add helper functions
`index_regular()`

and`index_summary()`

to help identify index variables

- remove
`feasts`

from dependencies as the functions required in`brolgar`

are actually in`fabletools`

. - add
`nearest_lgl`

and`nearest_qt_lgl`

- Gave more verbose names to the
`wages_ts`

data. - renamed
`sample_n_obs()`

to`sample_n_keys()`

and`sample_frac_keys()`

- renamed
`add_k_groups()`

to`stratify_keys()`

- removed many of the
`l_<summary>`

functions in favour of the`features`

approach. - rename
`l_summarise_fivenum`

to`l_summarise`

, and have an option to pass a list of functions. - rename
`l_n_obs()`

to`n_key_obs()`

- rename
`l_slope()`

to`key_slope()`

- added
`monotonic`

summaries and`feat_monotonic`

- rename
`l_summarise()`

to`keys_near()`

- make monotonic functions return FALSE if length == 1.
- add
`monotonic`

function, which returns TRUE if increasing or decreasing, and false otherwise. - re export
`as_tsibble()`

and`n_keys()`

from `tsibble - Data
`world_heights`

gains a continent column - Implement
`facet_strata()`

to create a random group of size`n_strata`

to put the data into (#32). Add support for`along`

, and`fun`

. - Implement
`facet_sample()`

to create facetted plots with a set number of keys inside each facet. (#32). `add_`

functions now return a`tsibble()`

(#49).- Fixed bug where
`stratify_keys()`

didn’t assign an equal number of keys per strata (#55) - Update
`wages_ts`

dataset to now just be`wages`

data, and remove previous`tibble()`

version of`wages`

(#39). - Add
`top_n`

argument to`keys_near`

to provide control over the number of observations near a stat that are returned. - change
`world_heights`

to`heights`

. - remove function
`n_key_obs()`

in favour of using`n_obs()`

(#62) - remove function
`filter_n_obs()`

in favour of cleaner workflow with`add_n_obs()`

(#63)

- Made brolgar integrate with
`tsibble`

.

- Added the
`world_heights`

dataset, which contains average male height in centimetres for many countries. #28 - created
`near_`

family of functions to find values near to a quantile or percentile. So far there are`near_quantile()`

,`near_middle()`

, and`near_between()`

(#11).`near_quantile()`

Specify some quantile and then find those values around it (within some specified tolerance).`near_middle()`

Specify some middle percentile value and find values within given percentiles.`near_between()`

Extract percentile values from a given percentile to another percentile.

- Create
`add_k_groups()`

(#20) to randomly split the data into groups to explore the data. - Add
`sample_n_obs()`

and`sample_frac_obs()`

(#19) to select a random group of ids. - Add
`filter_n_obs()`

to filter the data by the number of observations #15 - Remove unnecessary use of
`var`

, in`l_n_obs()`

, since it only needs information on the`id`

. Also gets a nice 5x speedup with simpler code - calculate all longnostics (#4)
- use the word
`longnostic`

instead of`lognostic`

(#9) `l_slope`

now returns`l_intercept`

and`l_slope`

instead of`intercept`

and`slope`

.`l_slope`

now takes bare variable names- Renamed
`l_d1`

to`l_diff`

and added a lag argument. This makes`l_diff`

more flexible and the function more clearly describes its purpose. - Rename
`l_length`

to`l_n_obs`

to more clearly indicate that this counts the number of observations. - Create
`longnostic`

function to create longnostic functions to package up reproduced code inside the`l_`

functions. - Added a
`NEWS.md`

file to track changes to the package.