Skip to content

Commit

Permalink
containers to struct
Browse files Browse the repository at this point in the history
  • Loading branch information
@cmaumet
cmaumet committed Jul 7, 2017
1 parent 729a523 commit 78dbe41
Show file tree
Hide file tree
Showing 3 changed files with 116 additions and 162 deletions.
124 changes: 49 additions & 75 deletions snpm_cp.m
View file
Original file line number Diff line number Diff line change
Expand Up @@ -217,22 +217,22 @@ function snpm_cp(CWD)
if strcmp(sDesFile, 'snpm_pi_OneSampT') || ...
strcmp(sDesFile, 'snpm_pi_ANOVAwithinS')
% Sign flipping
nidm_json('nidm_ErrorModel/nidm_hasErrorDistribution') = {'obo_nonparametricdistribution', 'obo_symmetricdistribution'};
nidm_json('nidm_ErrorModel/nidm_errorVarianceHomogeneous') = false;
nidm_json('nidm_ErrorModel/nidm_varianceMapWiseDependence') = 'nidm_IndependentParameter';
nidm_json('nidm_ErrorModel/nidm_hasErrorDependence') = 'nidm_IndependentError';
nidm.ErrorModel_hasErrorDistribution = {'obo_nonparametricdistribution', 'obo_symmetricdistribution'};
nidm.ErrorModel_errorVarianceHomogeneous = false;
nidm.ErrorModel_varianceMapWiseDependence = 'nidm_IndependentParameter';
nidm.ErrorModel_hasErrorDependence = 'nidm_IndependentError';
else
% Permutation
nidm_json('nidm_ErrorModel/nidm_hasErrorDistribution') = 'obo_nonparametricdistribution';
nidm_json('nidm_ErrorModel/nidm_errorVarianceHomogeneous') = true;
nidm_json('nidm_ErrorModel/nidm_varianceMapWiseDependence') = 'nidm_IndependentParameter';
nidm.ErrorModel_hasErrorDistribution = 'obo_nonparametricdistribution';
nidm.ErrorModel_errorVarianceHomogeneous = true;
nidm.ErrorModel_varianceMapWiseDependence = 'nidm_IndependentParameter';
% TODO: the 'obo_exchangeable' term is not yet in STATO
nidm_json('nidm_ErrorModel/nidm_hasErrorDependence') = 'obo_exchangeable';
nidm_json('nidm_ErrorModel/nidm_dependenceMapWiseDependence') = 'nidm_IndependentParameter';
nidm.ErrorModel_hasErrorDependence = 'obo_exchangeable';
nidm.ErrorModel_dependenceMapWiseDependence = 'nidm_IndependentParameter';
end

% TODO: check this is correct
nidm_json('nidm_ModelParameterEstimation/nidm_withEstimationMethod') = 'obo_ordinaryleastsquaresestimation';
nidm.ModelParameterEstimation_withEstimationMethod = 'obo_ordinaryleastsquaresestimation';

if isempty([H C])
error('SnPM:NoModel', 'No model specified; [H C] empty');
Expand All @@ -244,44 +244,38 @@ function snpm_cp(CWD)
error('SnPM:InvalidContrast','Contrast problem; wrong number of columns');
end

if size(CONT,1) > 1
contrastMap = containers.Map();
contrastMap('obo_contrastweightmatrix/prov:value') = CONT;
if size(CONT,1) > 1 % F-contrast
nidm.Contrasts(1).contrastweightmatrix_value = CONT;
warning('SnPM:FContrast', ...
'F contrast! F statistic images are being created.');
STAT = 'F';
contrastMap('nidm_StatisticMap/nidm_statisticType') = 'obo_Fstatistic';
nidm.Contrasts(1).StatisticMap_statisticType = 'obo_Fstatistic';

if (CONT(1,:) == -CONT(2,:))
CONT = CONT(1,:);
end
con_name = 'Positive';

nidm_json('Contrasts') = containers.Map({con_name}, {contrastMap});
nidm.Contrasts(1).StatisticMap_contrastName = con_name;
con_neg_name = '';
else
contrast_pos = containers.Map();
contrast_neg = containers.Map(); % There is no deep copy of
% containers.Map so we need a separate object from the start
con_name = 'Positive';
nidm.Contrasts(1).StatisticMap_contrastName = con_name;

con_neg_name = 'Negative';
nidm.Contrasts(2).StatisticMap_contrastName = con_neg_name;

STAT = 'T';
if bVarSm
contrast_pos('nidm_StatisticMap/nidm_statisticType') = 'nidm_smoothedtstatistic';
contrast_neg('nidm_StatisticMap/nidm_statisticType') = 'nidm_smoothedtstatistic';
nidm.Contrasts(1).StatisticMap_statisticType = 'nidm_smoothedtstatistic';
nidm.Contrasts(2).StatisticMap_statisticType = 'nidm_smoothedtstatistic';
else
contrast_pos('nidm_StatisticMap/nidm_statisticType') = 'obo_tstatistic';
contrast_neg('nidm_StatisticMap/nidm_statisticType') = 'obo_tstatistic';
nidm.Contrasts(1).StatisticMap_statisticType = 'obo_tstatistic';
nidm.Contrasts(2).StatisticMap_statisticType = 'obo_tstatistic';
end
contrast_pos('nidm_ContrastMap/nidm_contrastName') = ['Positive T-Contrast: [' mat2str(CONT) ']'];
contrast_pos('obo_contrastweightmatrix/prov:value') = CONT;
contrast_neg('nidm_ContrastMap/nidm_contrastName') = ['Negative T-Contrast: [' mat2str(-CONT) ']'];
contrast_neg('obo_contrastweightmatrix/prov:value') = -CONT;

% nidm_json('Contrasts').(con_name) = contrast_pos;
% nidm_json('Contrasts').(con_neg_name) = contrast_neg;

nidm_json('Contrasts') = containers.Map({con_name, con_neg_name}, ...
{contrast_pos, contrast_neg});
nidm.Contrasts(1).ContrastMap_contrastName = ['Positive T-Contrast: [' mat2str(CONT) ']'];
nidm.Contrasts(1).contrastweightmatrix_value = CONT;
nidm.Contrasts(2).ContrastMap_contrastName = ['Negative T-Contrast: [' mat2str(-CONT) ']'];
nidm.Contrasts(2).contrastweightmatrix_value = -CONT;
end

if rank(CONT)<size(CONT,1)
Expand Down Expand Up @@ -347,19 +341,13 @@ function snpm_cp(CWD)
df = q - r; %-Residual degrees of freedom
nPerm = size(PiCond,1); %-# permutations

nidm_json('nidm_NonParametricNullDistribution/nidm_numberOfPermutations') = nPerm;
nidm_json('nidm_NonParametricNullDistribution/nidm_hasResamplingScheme') = 'nidm_Permutation';
nidm_json('nidm_NonParametricNullDistribution/nidm_hasApproximationMethod') = 'nidm_MonteCarlo';
nidm_json('nidm_NonParametricNullDistribution/nidm_maximumNumberOfPermutations') = nPerm_max;
nidm.NonParametricNullDistribution_numberOfPermutations = nPerm;
nidm.NonParametricNullDistribution_hasResamplingScheme = 'nidm_Permutation';
nidm.NonParametricNullDistribution_hasApproximationMethod = 'nidm_MonteCarlo';
nidm.NonParametricNullDistribution_maximumNumberOfPermutations = nPerm_max;

con_names = nidm_json('Contrasts').keys;
for i = 1:numel(con_names)
contrast_map = nidm_json('Contrasts');
this_contrast = contrast_map(con_names{i});
this_contrast('nidm_StatisticMap/nidm_errorDegreesOfFreedom') = df;

contrast_map(con_names{i}) = this_contrast;
nidm_json('Contrasts') = contrast_map;
for i = 1:numel(nidm.Contrasts)
nidm.Contrasts(i).StatisticMap_errorDegreesOfFreedom = df;
end

%-Get ORIGIN, etc
Expand Down Expand Up @@ -389,8 +377,8 @@ function snpm_cp(CWD)
s_SnPM_save = [s_SnPM_save ' Vwt'];
bMask = 1;
end
% Add updated nidm_json to the saved variables
s_SnPM_save = [s_SnPM_save ' con_name con_neg_name nidm_json '];
% Add updated nidm to the saved variables
s_SnPM_save = [s_SnPM_save ' con_name con_neg_name nidm '];

%-Useful quantities - handy for later
%-----------------------------------------------------------------------
Expand Down Expand Up @@ -442,13 +430,14 @@ function snpm_cp(CWD)
descrip{ii}=sprintf('beta_%04d hats',ii);
Vbeta(ii)=snpm_clone_vol(Vt,fname{ii},descrip{ii});
end
nidm_json('ParameterEstimateMaps') = containers.Map(descrip, fname);
nidm.DesignMatrix_regressorNames = descrip;
nidm.ParameterEstimateMaps = fname;

Vbeta = spm_create_vol(Vbeta);

VResMS=snpm_clone_vol(Vt,'ResMS.img','Residual sum-of-squares');
VResMS=spm_create_vol(VResMS);
nidm_json('nidm_ResidualMeanSquaresMap/prov:atLocation') = 'ResMS.img';
nidm.ResidualMeanSquaresMap_atLocation = 'ResMS.img';

if bVarSm==0
str = sprintf('%c_{%d} statistic',STAT,df);
Expand All @@ -463,11 +452,11 @@ function snpm_cp(CWD)

Vmask=snpm_clone_vol(Vt,'mask.img',str);
Vmask=spm_create_vol(Vmask);
nidm_json('nidm_MaskMap/prov:atLocation') = 'mask.img';
nidm.MaskMap_atLocation = 'mask.img';

Vgmean=snpm_clone_vol(Vt,'GrandMean.img','GrandMean');
Vgmean=spm_create_vol(Vgmean);
nidm_json('nidm_GrandMeanMap/prov:atLocation') = 'GrandMean.img';
nidm.GrandMeanMap_atLocation = 'GrandMean.img';

if STAT=='T'
VT_pos=snpm_clone_vol(Vt,'snpmT+.img',[str,' (+ve)']);
Expand All @@ -479,45 +468,30 @@ function snpm_cp(CWD)
VCONSE=snpm_clone_vol(Vt,'conse.img', str);
VCONSE=spm_create_vol(VCONSE);

nidm_contrasts = nidm_json('Contrasts');
nidm_contrast = nidm_contrasts(con_name);

nidm_contrast('nidm_ContrastStandardErrorMap/prov:atLocation') = 'conse.img';

nidm_contrast('nidm_ContrastMap/prov:atLocation') = 'con+.img';
nidm_contrast('nidm_StatisticMap/prov:atLocation') = 'snpmT+.img';
nidm_contrasts(con_name) = nidm_contrast;
nidm.Contrasts(1).ContrastStandardErrorMap_atLocation = 'conse.img';
nidm.Contrasts(1).ContrastMap_atLocation = 'con+.img';
nidm.Contrasts(1).StatisticMap_atLocation = 'snpmT+.img';

% nidm_json('Contrasts').(con_name).('nidm_StatisticMap/prov:atLocation') = 'snpmT+.img';
VT_neg=snpm_clone_vol(Vt,'snpmT-.img',[str,' (-ve)']);
VT_neg=spm_create_vol(VT_neg);

VCON_neg=snpm_clone_vol(Vt,'con-.img',[str,' (+ve)']);
VCON_neg=spm_create_vol(VCON_neg);
% nidm_json('Contrasts').(con_neg_name).('nidm_StatisticMap__prov_atLocation') = 'snpmT-.img';
nidm_neg_contrast = nidm_contrasts(con_neg_name);
nidm_neg_contrast('nidm_StatisticMap/prov:atLocation') = 'snpmT-.img';
nidm_neg_contrast('nidm_ContrastMap/prov:atLocation') = 'con-.img';

% TODO: check it's fine to use same conse image for pos and neg contrasts
nidm_neg_contrast('nidm_ContrastStandardErrorMap/prov:atLocation') = 'conse.img';

nidm_contrasts(con_neg_name) = nidm_neg_contrast;

nidm_json('Contrasts') = nidm_contrasts;
nidm.Contrasts(2).ContrastStandardErrorMap_atLocation = 'conse.img';
nidm.Contrasts(2).ContrastMap_atLocation = 'con-.img';
nidm.Contrasts(2).StatisticMap_atLocation = 'snpmT-.img';
elseif STAT=='F'
VF=snpm_clone_vol(Vt,'snpmF.img',str);
VF=spm_create_vol(VF);

nidm_contrasts = nidm_json('Contrasts');
nidm_contrast = nidm_contrasts(con_name);
nidm_contrast('nidm_StatisticMap/prov:atLocation') = 'snpmF.img';
nidm_contrasts(con_name) = nidm_contrast;
nidm_json('Contrasts') = nidm_contrasts;
nidm.Contrasts(1).StatisticMap_atLocation = 'snpmF.img';

VFnum=snpm_clone_vol(Vt,'snpmFnum.img',str);
VFnum=spm_create_vol(VFnum);
nidm_contrast('nidm_ContrastExplainedMeanSquareMap/prov:atLocation') = 'snpmFnum.img';

nidm.Contrasts(1).ContrastExplainedMeanSquareMap_atLocation = 'snpmFnum.img';
end

VlP_pos=snpm_clone_vol(Vt, 'lP+.img', '-log10(uncor. non-para. P, +ve)');
Expand Down
Loading

0 comments on commit 78dbe41

Please sign in to comment.