Fixes skcsd figs (#60)

* Fix #43

* Fix indexing in skcsd figures

* First version of the skcsd complex morphology figure

* New version of figs

* Create two separate look-up tables for CSD and potential for sKCSD

The distance between source and estimation can be at least an order
of magnitude larger than the distance between the source and electrodes.
One interpolation table has either to be very large or miss vital
resolution.

* Fix lambdas and resolution of the interpolation table for figures

* Fixes for the article figure

Lower density of interpolation table.
Remove oscilatory injections to the soma.

* Fix interpolation tables for sKCSD

Some of the sKCSD calculations are performed in 2D space of two
variables on the morphology loop (source and estimation points positions
for potential), some are performed between 3D
cartesian variables (electrode positions) and 1D variables on the
morphology loop (source positions). Make separate spline interpolation routines
for both cases.

* Fix calculation of interpolation table for potentials

* Remove functions needed for sk_monaco, which is no longer used for skcsd

* Add transparency to plots

* Fixes for hexagonal electrode placement and Fig 8 complex morphology (insert hh to all soma)

* First version of the complex morphology figure

* Replication of Dori's figs

* Save number of columns and rows in figure names

* Fix electrode positions and drawing morphology

* Space after comma

* Scale lambdas

* Square electrode distribution, correct orientation

* Add more background to panels

* More transparent colormaps

* Allow for drawing morphologies and transforming CSD and pot in custom rectangles

* Remove unnecessary prints

* Figures 8 and 9

* Remove unnecessary import of run_LFP

* Fix writing morphology

* Indentation

* Move contruction of the table of sources and electrode positions to sKCSDcell for tests sake

* Add a function calculating euclidean distance for skcsd. If distance is smaller than 1 nm assume it is 1 nm

* Move construction of estimation tables to sKCSDcell to allow testing

* Fix typo

* Add electrode information to filenames for neuron simulations (for figure 10)

* Fix parameter choice for Y-cells

* Mainen morphology

* Same vmin and vmax

* Make same vmin and vmax

* More transparent morphologies in order to better see sKCSD maps

* Exact sKCSD caculation

* Exact sKCSD calculations (no interpolation)

* vmin i vmax for fig 3

* Make sure electrode will not cover sources on figure 9

* Complex morphology

* Add docstrings for some of the functions moved to sKCSDcell for the sake of the tests

* Add missing docstrings and remove some of the commented out code

* Remove test_get_src_ele_dists_1

* Remove unnecessary parameter from forward_model_3D execution

* Numpy 1.9 breaks Travis CI

* Revert "Numpy 1.9 breaks Travis CI"

This reverts commit 2df98ff.

* Reduce number of sources, interpolations and length of LFP to speed up tests.

* Remove illogical if

* Fix fname_base error

* Add docstrings for distance calculation and LoadData

* Increase numpy version to 1.9.0 to see if travis still breaks

* Remove Python3 LFPy provision

LFPy supports Python3.

* Add optional plotting morphology and electrode positions to CSD plots

Together with CSD or potential maps, plot electrode positions and
morphology optionally as well as grey polygons guiding the eye to
patches of higher activity.

* Grey shapes showing, where the activity is

* Small fixes in figs

* Make fig 5 compatible with eLife article

* Remove possibility of Monte Carlo integration

Remove Monte Carlo integration from kcsd and scikit-monaco from
dependencies. Remove obsolete skmonaco_available option from skcsd
figures.

* Remove scikit-monaco from additional packages in README

.travis.yml

@@ -18,7 +18,7 @@ env:
       NUMPY_VERSION="1.10.4" SCIPY_VERSION="0.16.1"
       SIX_VERSION="1.10.0" COVERAGE="true"
 
-    # # This environment tests conda extras skmonaco and matplotlib
+    # # This environment tests conda extras  matplotlib
     # - DISTRIB="conda_extra" PYTHON_VERSION="3.5"
     #   NUMPY_VERSION="1.10.4" SCIPY_VERSION="0.16.1"
     #   SIX_VERSION="1.10.0" COVERAGE="true"
@@ -35,7 +35,7 @@ env:
     # - DISTRIB="ubuntu" PYTHON_VERSION="2.7"
     #  NUMPY_VERSION="1.8.0" SCIPY_VERSION="0.14.0" SIX_VERSION="1.10.0"
     #  COVERAGE="true"
-
+    
     # - DISTRUB="ubuntu_extra" PYTHON_VERSION="2.7"
     #   SKMONACO_VERSION="0.2" NUMPY_VERSION="1.13.0" SCIPY_VERSION="1.0.0"
     #   MATPLOTLIB_VERSION="0.99"  SIX_VERSION="1.10.0" COVERAGE="true"

README.rst

@@ -75,7 +75,6 @@ Requirements
 Additional Packages
 -------------------
 
-- scikit-monaco 0.2 (for monte carlo type integration for 3D case)
 - matplotlib 0.99 (for visualization and plotting)
 
 
@@ -133,7 +132,6 @@ If you use this software in published research please cite the following work
 
 5) Ness, T. V., Chintaluri, C., Potworowski, J., Łeski, S., Głabska, H., Wójcik, D. K. & Einevoll, G. T. (2015) 'Modelling and Analysis of Electrical Potentials Recorded in Microelectrode Arrays (MEAs).' Neuroinformatics 13(4), 403-426.
 
-
 6) Cserpan, D., Meszena, D., Wittner, L., Toth, K., Ulbert, I., Somogyvari, Z., Wójcik, D. K. (2017) 'Revealing The Distribution Of Transmembrane Currents Along The Dendritic Tree Of A Neuron From Extracellular Recordings.' eLife (2017) 6:e29384