/
TODO
163 lines (162 loc) · 7.51 KB
/
TODO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
TODO
--------------------
- make latest SVN/GIT version available from somewhere!
- cleanup old files in GIT history (there are some abf files somewhere! r1190-1199)
- make function to save text file from trace!
- website options:
- keep INCF and add more wikis. pros: professional place. cons: wiki files not in repo.
- switch to github: wikis are in different repo, but versioned, nice
web page support for projects.
- make entry in Mathworks page. pros: associated with author name
- best solution: keep all three, make github main place with web
pages for all three projects: pandora, cgmplots, param_fitter.
Keep everything in github, but push to SVN on INCF? NeuroTools
keeps everything external.
- add support for neuroshare matlab import [isn't this done?]
- change names of major classes to make more sense
- modify license to include attribution in scientific publications
- add test suite?
- add more examples to lyx documentation. [doing wikis on incf instead]
- replace exist functions with a 'var' 2nd param [what? No!]
- add pclamp import function abf_atf_import from Matlab file exchange
in the utils/ folder. [abf2load works now?]
- make trace accept atf, abf and Spike2 formats
- use the neuroshare input specification?
- put license, add SQL export function, and publish package
- add neurosage hdf5 support
- ask to correct link on http://home.earthlink.net/~perlewitz/sftwr.html
comes first in google when searched for "neuroscience software".
also link to http://scholarpedia.org/article/Encyclopedia_of_Computational_Neuroscience
- rename plot_abstract, joined_db to more meaningful names.
- generalize histogram_db to hold 2 and 3-dimensional histograms
- open tools/utils dir and move needed stuff from cen_matlab and alftoolbox into there.
- make a paper production class capable of generating figures for paper automatically
- make plan and requisites for version procession
- put ItemIndex as param
- improve visualization of cip_traces
- create custom class to eliminate high-variance neurons? how to get sdb?
- plot matches that vary a lot in parameter values, but same outcome
- move multi_fileset_gpsim_cns2005 into djlab
+ move plotfICurve to model_ct_bundle
- make comparisonReport use reportCompareModelToPhysiolNeuron
+ for version 0.6:
- add new physiology dataset loader
+ for version 0.7:
+ new AP thr method, and look at all spikes
+ also look at phys spikes, needs some smoothing
+ rename displayRow to displayRows
+ fix SE calc
- for version 0.8:
+ move props to bottom of params_tests_dataset and tests_db
+ remove redundant props from subclasses
- for version 1.0:
+ add dataset connection to tests_db [created the bundle concept]
- make version-aware loader?
X put it in props for now?
- put params in dataset, params_results_profile [?]
- fix fillederrorbar with new Matlab
- Make tests_db save and load its dataset on request
- overload saveobj and loadobj methods
- tests_db
+ add arithmetic operations that operate on the data matrix [can add columns]
+ add horzcat, vertcat methods to concat rows or columns (!)
+ print column numbers in display
+ define all operations as row logical functions
+ complex subsref for tests_db.
- add plot-frame and docgen classes
- plot_bars puts too much x-margin space
- move plot_bars to utils/?
- make plot_superpose return correct errorbar extents
- allow adding insets to plots
- clustering histograms?
+ fix varboxplot to return correct y-axis limits
+ add a PaperPosition property to figures to make printable figures.
+ allow fontsize change to fit more on a single page [can use export size for that]
+ Replace older plot methods with plotFigure(plot_abstract(.)) combinations
- what to do with Interpreter=None stuff?
+ matrix color plot
+ allow adjusting location of title in stack plots
+ invar param value should be indicated on plotPages
+ classes should have function that returns plot_abstract
+ @plot_abstract is one x-y plot w/ legends, etc.
+ @plot_stack, places plots on subplots.
+ sortrows in histograms.m
+ histograms should have same axis limits
+ @plot_abstract/superposePlots, superposes plots on same axis, and
return a new plot_abstract object
- objects don't take more space when they're used again, has implications on design:
- put links to previous db's for a chain of operations, but they must not be saved!
+ put link to dataset from db [made bundles]
- comparing physiology data
- improve convenience by adding custom functions to bundle classes
- cluster params of best matches
- collect params of all best matches and cluster
+ allow easily choosing subset of measures
+ made color-coded error distance display
+ make tabular display
- Make params_tests_profile to use arrays rather than cell arrays
- do for rest of profile elements
+ had to remove size member function
+ may need to modify subsref [if an array is passed, then call built-in function]
+ then make plot methods operate on arrays and return plot arrays
- update documentation
- perl doclet generator
- make first description no longer than one line on all constructors
- make sure an example accompanies each function
- Make PDF with code design
- add class diagram with converter methods
- add diagram for bundles
+ update README
- high-order methods:
+ factor analysis
+ shuffle predicition
- pca
- ica
- clustering (use for finding multiple solutions?)
- need this for finding parameter clusters matching phys data?
- no, use hierarchical clustering to find number of clusters and centers
- use PCA to display clustering results
- use ANOVA for showing significance of separation
+ make dualcip a new db and make it load profiles, etc. [made multi_fileset]
+ add method in dataset to load item by reading database row or in db vice-versa.
- use db id for the prof id
+ ignore all parameters in comparisons?
+ plot average neuron comparison matches.
+ remove DAHPMag from depol_tests
+ remove not-a-spikes from spikes object
+ analyze each spike individually, then generate mean, mode, median, and unimodality score
+ add an addItem method to datasets
+ new subclass for custom analysis
- add plotTestStats function to group things together [?]
- CVS
- present CVS, start using it
- move this repository to djlab account?
+ add min ahp time to DB
+ replace '_' in latex outputs
+ ignore NeuronId while comparison and do the comparison again
+ profile:
+ make a converter method from params_tests_db to get the profile
+ make the profile constructor simple
+ put all test results into the results structure
+ overload the get method to look into the results structure
+ matrix stack plot
+ new plot & data type for mean and stds of each test and param value
+ swapRowsPages, then find statsMeanStd
+ statsMeanStd needs page-capable mean and std funcs
+ create special plot with vectors and errorbars
+ do it over all parameters in the profile
+ then plot a test for all parameters in the same way
+ use boxplots
+ poster size: 1.75x1.1 m^2, 68"x43"
+ sortrows(tests_db) only sorts one page and applies to all?
+ invarValues should've been sorted already? [no, only other columns are sorted]
+ add calcSlowInact to @cip_trace
+ add cip_trace_fileset class
+ make abstract profile class to describe the getResults method
+ Make Matlab contents file
+ getResults for @spikes, @trace
+ Make README file
+ make spike checks in cip_trace_profile
+ make specialized directory for brute_aug_04, move cip_trace_profile there
+ first make a general cip_trace_profile which takes ready-made objs
+ remove conflicting classes from the cen_matlab module