-
Notifications
You must be signed in to change notification settings - Fork 1
/
process.m
134 lines (108 loc) · 5.69 KB
/
process.m
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
% Generate the model covariates. I have custom scripts
% that form the model covariates from the experimental data. It generally
% involves filtering the experimental data with various types of smooth
% functions. My ecosystem allows the user to choose which experimental
% variables to filter, which functions to filter with, and other parameters.
% The "define_filters" functions generate a structure with this information
% which is later used by the function that actually does the filtering,
% filter_and_insert.m
filt_struct = TW_define_filters(0,model_num, dt); % For external covariates
filt_struct2 = TW_define_filters2( num_nrn , num_spk_history_bf ,0, dt); % For spike history covariates
% Bookkeeping: Get number of bins per trial
bins_per_trial = cellfun(@size,movement_data,'UniformOutput',0);
bins_per_trial = cell2mat(bins_per_trial);
bins_per_trial = bins_per_trial(:,1); % for the non-padded cell array
bins_per_trial2 = reshape([bins_per_trial bins_per_trial]',[],1); % for the padded cell array
% Filter neural data for spike history terms
if include_spk_history
% Add padding between trials to prevent spillover
temp = cell2mat(spikes);
spikes_to_filter = mat2cell(temp,bins_per_trial,num_nrn);
spikes_to_filter = [spikes_to_filter cellfun(@(x)x*0,spikes_to_filter,'UniformOutput',0)]; % pad
spikes_to_filter = cell2mat(reshape(spikes_to_filter',[],1)); % reshape
% Filter
spk_hist_cov = full(filter_and_insert( spikes_to_filter, filt_struct2 ));
spk_hist_cov = mat2cell(spk_hist_cov,bins_per_trial2,num_nrn*num_spk_history_bf);
spk_hist_cov = spk_hist_cov(1:2:end);
% Trim
spk_hist_cov = trim_edges(spk_hist_cov,trim_times,dt,2,movement_data);
spk_hist_cov = cell2mat(spk_hist_cov);
end
% Filter external covariates
if model_num == 8 || model_num == 9
% Make duplicate with 0s for padding (every other "trial" is
% blank). This prevents filter spillover into adjacent trials.
temp = [movement_data cellfun(@(x)x*0,movement_data,'UniformOutput',0)];
% Reshape
temp2 = cell2mat(reshape(temp',[],1));
% Filter
X_temp = full(filter_and_insert(temp2,filt_struct));
% Remove padding
X_cell_temp = mat2cell(X_temp,bins_per_trial2,size(X_temp,2));
X_cell_temp2 = X_cell_temp(1:2:end);
% Trim edges
X_cell_temp2 = trim_edges(X_cell_temp2,trim_times,dt,2,movement_data);
X_cell(:,1,1) = X_cell_temp2;
clear X_cell_temp
elseif model_num == 10
% Make a cell for speed (not included in original kinematic data)
kin_all = cell2mat(kinematics_data(trials));
speed_temp = sqrt(kin_all(:,3).^2 + kin_all(:,4).^2);
speed_cell = mat2cell(speed_temp,bins_per_trial,1);
% Gather all covariates to be filtered into a single cell array
cov_to_filter = [cell2mat(movement_data) cell2mat(kinematics_data(trials)) cell2mat(speed_cell(trials))];
cov_to_filter(:,4:5) = []; % Get rid of x,y-position
cov_to_filter2 = mat2cell(cov_to_filter,bins_per_trial,size(cov_to_filter,2));
% Make duplicate with 0s for padding (every other "trial" is
% blank). This prevents filter spillover into adjacent trials.
temp = [cov_to_filter2 cellfun(@(x)x*0,cov_to_filter2,'UniformOutput',0)];
% Reshape
temp2 = cell2mat(reshape(temp',[],1));
% Filter
X_temp = full(filter_and_insert(temp2,filt_struct));
% Remove padding
X_cell_temp = mat2cell(X_temp,bins_per_trial2,size(X_temp,2));
X_cell_temp2 = X_cell_temp(1:2:end);
% Trim
X_cell_temp2 = trim_edges(X_cell_temp2,trim_times,dt,2,movement_data);
X_cell(:,1,1) = X_cell_temp2;
clear X_cell_temp
elseif model_num == 11
% Make a cell for speed (not included in original kinematic data)
kin_all = cell2mat(kinematics_data(trials));
speed_temp = sqrt(kin_all(:,3).^2 + kin_all(:,4).^2);
speed_cell = mat2cell(speed_temp,bins_per_trial,1);
% Gather all covariates to be filtered into a single cell array
cov_to_filter = [cell2mat(movement_data(trials)) cell2mat(kinematics_data(trials)) cell2mat(speed_cell(trials))];
cov_to_filter(:,4:5) = []; % Get rid of x,y-position
cov_to_filter2 = mat2cell(cov_to_filter,bins_per_trial,size(cov_to_filter,2));
% Make duplicate with 0s for padding (every other "trial" is
% blank). This prevents filter spillover into adjacent trials.
temp = [cov_to_filter2 cellfun(@(x)x*0,cov_to_filter2,'UniformOutput',0)];
% Reshape
temp2 = cell2mat(reshape(temp',[],1));
% Filter
X_temp = full(filter_and_insert(temp2,filt_struct));
% Remove padding
X_cell_temp = mat2cell(X_temp,bins_per_trial2,size(X_temp,2));
X_cell_temp2 = X_cell_temp(1:2:end);
% Trim
X_cell_temp2 = trim_edges(X_cell_temp2,trim_times,dt,2,movement_data);
X_cell(:,1,1) = X_cell_temp2;
clear X_cell_temp
end
% Trim edges for spikes
spikes2 = spikes;
spikes = trim_edges(spikes,trim_times,dt,2,movement_data);
spikes_warped(:,:,1) = spikes;
% After trimming, get new number of bins per trial
bins_per_trial = cellfun(@size,X_cell,'UniformOutput',0);
bins_per_trial = cell2mat(bins_per_trial);
bins_per_trial = bins_per_trial(:,1);
% Other pre-processing
X_cell(:,2:num_WV,1) = repmat(X_cell(:,1,1),[1 (num_WV-1)]); % Duplicate starting X matrix across WV_folds
% Set up some spike history stuff
num_exp_cov = size(cell2mat(X_cell(:,1,1)),2);
for nrn = 1:num_nrn
idx_spk_hist_cov{nrn} = num_spk_history_bf*(nrn - 1) + [1:num_spk_history_bf] + num_exp_cov;
end