NEWS.md
keys_near
related to factorsfeat_diff_summary()
functions to help summarise diff(). Useful for exploring the time gaps in the index
. (#100)facet_sample()
now has a default of 3 per plotnear_quantile()
, the tol
argument now defaults to 0.01.tbl_ts
objects for keys_near()
- #76pisa
containing a short summary of the PISA dataset from https://github.com/ropenscilabs/learningtower for three (of 99) countriesindex_regular()
and index_summary()
to help identify index variablesfeasts
from dependencies as the functions required in brolgar
are actually in fabletools
.nearest_lgl
and nearest_qt_lgl
wages_ts
data.sample_n_obs()
to sample_n_keys()
and sample_frac_keys()
add_k_groups()
to stratify_keys()
l_<summary>
functions in favour of the features
approach.l_summarise_fivenum
to l_summarise
, and have an option to pass a list of functions.l_n_obs()
to n_key_obs()
l_slope()
to key_slope()
monotonic
summaries and feat_monotonic
l_summarise()
to keys_near()
monotonic
function, which returns TRUE if increasing or decreasing, and false otherwise.as_tsibble()
and n_keys()
from `tsibbleworld_heights
gains a continent columnfacet_strata()
to create a random group of size n_strata
to put the data into (#32). Add support for along
, and fun
.facet_sample()
to create facetted plots with a set number of keys inside each facet. (#32).add_
functions now return a tsibble()
(#49).stratify_keys()
didn’t assign an equal number of keys per strata (#55)wages_ts
dataset to now just be wages
data, and remove previous tibble()
version of wages
(#39).top_n
argument to keys_near
to provide control over the number of observations near a stat that are returned.world_heights
to heights
.n_key_obs()
in favour of using n_obs()
(#62)filter_n_obs()
in favour of cleaner workflow with add_n_obs()
(#63)world_heights
dataset, which contains average male height in centimetres for many countries. #28near_
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.add_k_groups()
(#20) to randomly split the data into groups to explore the data.sample_n_obs()
and sample_frac_obs()
(#19) to select a random group of ids.filter_n_obs()
to filter the data by the number of observations #15var
, in l_n_obs()
, since it only needs information on the id
. Also gets a nice 5x speedup with simpler codelongnostic
instead of lognostic
(#9)l_slope
now returns l_intercept
and l_slope
instead of intercept
and slope
.l_slope
now takes bare variable namesl_d1
to l_diff
and added a lag argument. This makes l_diff
more flexible and the function more clearly describes its purpose.l_length
to l_n_obs
to more clearly indicate that this counts the number of observations.longnostic
function to create longnostic functions to package up reproduced code inside the l_
functions.NEWS.md
file to track changes to the package.