Changelog

Version 0.1.1

Changes

• examples is not be treated as a module anymore (as it should). To use the examples, one should go to the example folder to run the scripts.

• Added line creation in get_colors_and hatches_lists, renamed it get_colors_hatches_lines_lists and used it in more places.

Added

• Example with a cell, script to record currents/ionic concentrations from it, and script to plot its currentscape.

• Test to run the new example, running the cell and plotting its currentscape.

Fixed

• Fixed legend for currentscape with no hatches

• Removed minor yticks. They were popping out when not using autoscale for ylim and ticks with log scale

• Dotted horizontal lines at tick level. They were not the right length when data with adaptable timesteps were given

Removed

• extract_currs module

• unused .gitreview file

• __init__.py file in examples. examples folder should not be treated as a module.

Version 0.1.0

Changes

• use fill_between instead of imshow or barplot (more precise, less pixelised and consumes less memory)

• default extension for figure saving in config is now pdf (in order to not loose details)

Added

• new ‘legacy_method’ entry in config for currents and currentscape to access legacy imshow and barplots

Version 0.0.12

Changes

• Turned some print statements into logging statements

• Refactored Main Protocol’s run function in extract_currs

• Refactored output writing into a write_output function in extract_currs

• Turned os.path statements into pathlib.Path statements

• Refactored extract_currs/features/define_efeatures

• Turned string formatting into f-strings

• Refactored the loading of release_params into a function

• Moved bluepyopt, neuron and scipy dependencies as extras in setup.py

• Refactored currents.reorder_currents_and_names for better quality code

• Used np.asarray instead of np.array when possible

• Refactored color and pattern mapping into functions

• Renamed col_idxs into curr_idxs in several function arguments

Added

• Unit tests for all currentscape functions that do not use matplotlib

• Linting checks to the extract_currs module

• Test for the new write_output function

• Installation instructions in the README

Removed

• Unused logging.basicConfig

• Print statements

• extract_currs/YeoJohnson.py

• extract_currs/features/eFELFeatureExtra

• extract_currs/features/SingletonWeightObjective

• Non-necessary stuff from extract_currs/features/define_efeatures

• Score calculations during Main Protocol

• Non-necessary variables form extract_currs’ create_protocols

• __init__ function in SweepProtocolCustom that was the same as the parent class

Fixed

• Tox

• Barplot shrinking when chunk size has to be resized

Version 0.0.11

Changes

• Improved memory use by deleting large variables when they are not of use anymore.

• Improved memory use by using np.int8 in create_black_line function.

• Made sure that every data in Datasets are numpy arrays.

Version 0.0.10

Changes

• Forced fixed timesteps in extract_currs, since currentscape expects data with fixed timesteps.

Version 0.0.9

Added

• New SAHP stimulus for current extraction.

Version 0.0.8

Fixed

• Test data in test_currentscape_functions.

Version 0.0.7

Fixed

• Unit conversion when turning current density into currents in extract_currs.

Added

• Added tests for data processing functions.

• Added warning when there is not enough colors in colormap.

Changes

• Refactored currentscape into smaller, more meaningful modules.

• Changed default left-side adjusting,

  so that the labels and ticks do not overlap.

• Changed dict key for blackline thickness and x-chunksize in config

  since they should not belong exclusively to currentscape anymore.

• Changed ‘labels’ and ‘ticklabels’ to ‘ylabels’ and ‘yticklabels’ in config[“show”]

  to distinguish them from ‘xlabels’ and ‘xticklabels’.

New Features

• Reorder the currents in each subplot,

  so that the currents with the largest contribution lay at the top. The legend is reordered accordingly.

• Added an autoscaling for y limits and ticks for current (and ions) subplots.

• Added a new subplot to currentscape: absolute currents.

• Absolute currents can be displayed as lines or as a stackplot.

• Allowed extract_currs to extract ionic concentration.

• Added a new subplot to currentscape: ionic concentration.

• Added a new subplot: pie charts showing the overall contribution of currents.

• Label and ticklabel of the x axis can now be displayed on the bottom plot.

• Gridlines corresponding to the x ticks can be displayed on all the plots.

• x ticks can be custom or generated automatically.

• Currentscape plotting can be disabled in the config.

Version 0.0.6

Fixes

• Currentscape used to fill empty data (e.g. no inward/outward current at all)

  with the first color in data. Now it fills it with white.

• Fixed config example in README and doc.

• Removed top and bottom frame of currentscape which was hiding part of the data.

• Fixed central black line separating currentscape plots that weas hiding part of the data.

New Features

• Added new colormap choices.

• Added possibility of putting patterns (hatches) on top of color palette.

• When using patterns, matplotlib uses bar plots, which takes some time to compute,

  so reducing x-resolution was enabled.

• Chunksize (when reducing x-resolution) is corrected when not a divisor of data size.

• Patterns color, design and ‘density’ can be customised.

• The legend y position can be customised.

• Added a ‘mixer’ that prevent two successive currents to have the same color or pattern.

• Added library to easily extract currents from any location (soma or other).

• Now uses neuron in extract_currs lib to get the segment area and use it to turn

  current densities into currents.

• Can use default protocols from singlecell-optimization, or custom protocols.

• Added tests for extract currents.

Version 0.0.4

• Added docs to tox envlist.

Version 0.0.3

Fixes

• Fixed non pylint-conform style.

Version 0.0.2

New Features

• Added documentation.

Version 0.0.1

New Features

• Currentscapes can be plotted using code from Alonso and Marder (2019).

• Labels, ticks and legend are automatically added to the figure.

• Most of the style is hard-coded, to stay close to the original paper,

  but the figure display can still be adapted to most data, using a configuration file.

• Added example from original paper.

• Added example in README.md with link to homemade data.