All notable changes to this project will be documented in this file.
This project adheres to Semantic Versioning.
- Include ipynb files in tests, and when measuring test coverage
- Changed the
radius
argument of a Circle (in Maps) toarea
.
- Fixes deprecation warnings with matplotlib's warn parameter.
- Small cleanups: update to latest version of folium.
- Add some additional documentation.
- Fixes bug with
Table#hist
.
- Adds support for NumPy v1.18.0+.
- Fixes multiple bugs with the
Table#remove
.
- Fixes bug with grouping tables with np.float64('nan') type objects.
- Fixed a bug that was introduced by v0.15.1.
- Adds more flexibility to
Marker#map_table
for user-defined options for markers.
- Fixed a bug related to histogram shading
- Added support for shading part of a histogram, e.g., for highlighting the tail of a distribution.
- Adds optional argument to
Table#from_df
that keeps the index when converting from a Pandas dataframe.
- Declares all dependencies required for this package in requirements.txt.
- Adds a warning to help students realize why
Table.with_columns(...)
doesn't work.
- Adds support for other built-in tile sets other than
OpenStreetMap
toMap
.
- Adds support to
Map#read_geojson
for reading in GeoJSON from links.
make_array
automatically chooses int64 data type instead of int32 on Windows.
- Changes default formatting of numbers in printed output to 12345 instead of 12,345.
- Allows for the following notations ("floating arguments") to be used with
Table#take
andTable#exclude
: ex.t.take(0, 1, 2, 3)
andt.exclude(0, 2, 4)
.
- Removes deprecated argument for
Table#__init__
.
- Update mapping code to work with the latest version of Folium (0.9.1).
- Changes
Table#scatter
's argument name ofcolors
togroup
to mirrorTable#hist
. - Makes a grouped scatterplot's legend identical to a group histogram's legend.
- Fixes bug where x-label doesn't show up for grouped histogram in certain conditions.
- Fixed bug where Table#hist was sometimes truncating the x-axis label.
- Fixes bug where error terms show up while plotting
- Fixes bug where joining tables that have columns that are already duplicated will sometimes join incorrectly.
- Fix bug where we warned inappropriately when passing a string to an
are.*
predicate.
- Switch from pandas.read_table to pandas.read_csv, to avoid deprecation warnings. Shouldn't change the behavior of the library.
Table.append_column
now returns the table it is modifying.
- Add
shuffle
function toTable
.
- Added
join
for multiple columns.
- Allow NumPy arrays to be appended into tables.
- Added optional formatters to "Table.with_column", "Table.with_columns", and "Table.append_column".
- Warning added for comparing iterables using predicates incorrectly.
- 'move_column' added.
- Created new methods 'first' and 'last'.
- 'append_column' now returns the table it is modifying.
- 'move_to_end' and 'move_to_start' can now take integer labels.
- Fixes test suite and removes all deprecated code in the test suite caused by deprecated API calls from the datascience library.
- Adds
hist_of_counts
function
- Fixes minor issues introduced by matplotlib 2.x upgrade (#315)
- Fixes a bug in HTML table generation (#315)
- Add
sample_proportions
function.
- Fix
OrderedDict
bug inTable.hist
.
- Fix
CurrencyFormatter
to handle commas. - Fix
Table.hist
to keep histograms in the order of the columns.
- Fix
join
so that it keeps all rows in the inner join of two tables.
- Added
group_barh
andgroup_bar
to plot counts by a grouping category, a common use case. - Added options to
hist
to produce a histogram for each group on a column. - Deprecated Table method
pivot_hist
. Added an option tohist
to simulatepivot_hist
's behavior.
- DistributionFormatter added.
- Fix bug for relabeled columns that had a format already.
- Circles bound to values determine the circle area, not radius.
- Scatter diagrams can take data-driven size and color parameters.
- Changed signature of
apply
,hist
, andbin
to accept multiple columns without a list - Deprecate
hist
argument namecounts
in favor ofbin_column
- Rename various positional args (technically could break some code, but won't)
- Unified
with_column
andwith_columns
(not a breaking change) - Unified
group
andgroups
(not a breaking change)
- Added "Table.remove"
- Added
proportions_from_distribution
method todatascience.util
. (993e3d2) Table.column
now throws a descriptiveValueError
instead of aKeyError
when the column isn't in the table. (ef8b319)
Breaking changes
- Change default behavior of
table.sample
towith_replacement=True
instead ofFalse
. (3717b67)
Additions
- Added
Map.copy
. - Added
Map.overlay
which overlays a feature(s) on a new copy of Map. (315bb63e)
- Remove rogue print from
table.hist
- Added predicates for string comparison:
containing
andcontained_in
. (#231)