Changelog

v2.0.0 2021-08-25 - see What’s new in Version 2

v1.6.6 2021-07-06

• workaround for groupby bug in pandas 1.3.0 (GH42395)

v1.6.5 2021-06-14

• bugfix for staircase.Stairs.mode()

v1.6.4 2021-03-16

• bugfix for broken functionality for multiplication, or division, of staircase.Stairs with a number

v1.6.3 2021-02-26

• bugfix for missing lower_how and upper_how parameters in staircase.Stairs.resample()

• renamed Stairs._values to avoid error when performing Series.groupby.sum in pandas>=1.2

v1.6.2 2020-01-13

• efficiency improvement for staircase.Stairs.multiply()

v1.6.1 2020-12-30

• bugfix for staircase.Stairs.plot() when using datetimes and step function has no step changes

v1.6.0 2020-11-10

• support for timezones added

• bugfix for staircase.sample() when points == None (-inf is no longer included)

• bugfix for staircase.aggregate() where Stairs objects have non-zero value at -inf

v1.5.2 2020-10-20

• bugfix for staircase.Stairs.layer() when None appears in vector parameters start and end

v1.5.1 2020-10-15

• removed dependency on ‘private’ methods in sortedcontainers

• increased upper limit for version dependency on sortedcontainers

v1.5.0 2020-10-12

• fixed typo in diff docstring

• extended staircase.Stairs.values_in_range() to allow specification of how endpoints of domain should be evaluated

• extended staircase.Stairs.min() to allow specification of how endpoints of domain should be evaluated

• extended staircase.Stairs.max() to allow specification of how endpoints of domain should be evaluated

• extended staircase.Stairs.sample() to allow specification of how endpoints of domain should be evaluated

• extended staircase.Stairs.resample() to allow specification of how endpoints of domain should be evaluated

• added staircase.Stairs.rolling_mean()

• added example usage of rolling_mean to Queue Analysis tutorial

v1.4.0 2020-10-05

• extended staircase.Stairs.corr() to facilitate cross-correlation and autocorrelation

• extended staircase.Stairs.cov() to facilitate cross-covariance and autocovariance

v1.3.0 2020-10-01

• added staircase.Stairs.describe()

• added staircase.Stairs.var() (variance)

• added staircase.Stairs.std() (standard deviation)

• added staircase.Stairs.corr() (correlation)

• added staircase.Stairs.cov() (covariance)

• added staircase.cov() (pairwise covariance matrix)

• added staircase.corr() (pairwise correlation matrix)

v1.2.0 2020-09-23

• added staircase.Stairs.hist()

• added staircase.Stairs.ecdf_stairs()

• added staircase.hist_from_ecdf()

• added staircase.make_test_data()

• staircase.Stairs.percentile_Stairs() pending deprecation in favour of staircase.Stairs.percentile_stairs()

v1.1.1 2020-09-22

• bugfix for staircase.Stairs.sample() when parameter x = float(“-inf”)

v1.1.0 2020-09-15

• added staircase.Stairs.shift()

• added staircase.Stairs.diff()

v1.0.3 2020-09-08

• deep parameter added to staircase.Stairs.copy() method for pandas compatability

v1.0.[1|2] 2020-09-02

• bugfix with Pandas 1.1.x where SortedSet cannot be used as basis for Series or DataFrame

v1.0.0 2020-09-01

• updated documentation to include A note on interval endpoints

• parameter start in staircase.Stairs.layer() made optional to make method symmetric with respect to time

• removed staircase.Stairs.evaluate method (superseded by staircase.Stairs.sample())

Speed comparison - V1 vs V2 (using dates)