Skip to content

Commit

Permalink
CDC updates OR
Browse files Browse the repository at this point in the history
  • Loading branch information
@kruggles7
kruggles7 committed Nov 14, 2014
1 parent 2be001f commit efd1762
Show file tree
Hide file tree
Showing 27 changed files with 4,757 additions and 18 deletions.
621 changes: 621 additions & 0 deletions heatmaps/demographics_2013_CI.m
View file

Large diffs are not rendered by default.

45 changes: 45 additions & 0 deletions heatmaps/run_demographics_2013.m
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,45 @@
cd ..
cd ..
cd data
cd Controls_061514
sex=importdata('sex-NaN.txt', '\t');
race=importdata('race-NaN.txt', '\t');
weight=importdata('weights-NaN.txt','\t');
grade=importdata('grade-NaN.txt','\t');
cd ..
cd results_103114
cd cat
ques=input ('Enter in the question number you want to use (ex. Q01): ', 's');
question_mat=importdata([ques '-cat-NaN.txt'], '\t');
[r,c]=size(question_mat);
labels=question_mat(:,1);
question_mat=question_mat(:,2:c);
years=[2001 2003 2005 2007 2009 2011 2013];
start_year=0;
end_year=0;
years2=double.empty;
count=1;
for i=1:numel(years)
indx=find(labels==years(i));
if numel(indx)==1
if start_year==0
start_year=i;
else
end_year=i;
end
years2(count)=years(i);
count=count+1;
end
end

[r,c]=size(sex);
weight=weight(start_year:end_year,2:c);
race=race(start_year:end_year,2:c);
sex=sex(start_year:end_year,2:c);
grade=grade(start_year:end_year,2:c);
cd ..
cd ..
cd ..
cd programs
cd heatmaps
[conf_mat, total, x_mat, n_mat, conf_mat2] = demographics_2013_CI( question_mat, race, sex, grade, weight, years2 );
154 changes: 154 additions & 0 deletions odds_ratio/OR_2013_categorical.m
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,154 @@

function [] = OR_2013_categorical(ques1)

k=1; %counter for rows in rel_risk_cell and odds_ratio_cell

load reverse_code_091914
cd results_103114
cd cat
file1=[ques1 '-cat-NaN.txt'];
quest_1=importdata(file1, '\t');
filename1=ques1;
a1=char(ques1);
ct='';
for q=2:numel(a1)
c1=a1(q) ;
ct=[ct c1];
end
ct=str2num(ct);
q1_RC=reverse_code(ct,1) ;
[r,c]=size(quest_1);
max_q1=max(max(quest_1(:,2:c)));

%filename2
files2=dir(fullfile('/ifs/data/proteomics/projects/cdc/matlab/results_103114/NaN/', '*.txt'));
cd ..
cd NaN
N=length(files2);
for CAT=1:max_q1 %for each categorical response for question 1
for n=1:N %for each question 2
quest_2=importdata(files2(n).name, '\t');
filename2='';
a2=char(files2(n).name);
b2=strfind(a2,'-');
for m=1:(b2(1)-1)
c2=a2(m);
filename2=[filename2 c2];
end
ct=[];
for m=2:(b2(1)-1)
c2=a2(m);
ct=[ct c2];
end
files2(n).name;
ct=str2num(ct);
q2_RC=reverse_code(ct,1) ;

%PROCESS THE COMBO
if ( isnan(q2_RC)==0 && isnan(q1_RC)==0)

[r1,c1]=size(quest_1);
[r2,c2]=size(quest_2) ;
year=[2001 2003 2005 2007 2009 2011 2013];
year1=quest_1(:,1);
year2=quest_2(:,1);
year_final=double.empty;
quest_1F=double.empty;
quest_2F=double.empty;
counter=1;
s=0; %start for 1
e=1; %start for 2
for ii=1:numel(year)
y=year(ii);
indx1=find(year1==y);
indx2=find(year2==y);
if numel(indx1)>0 && numel(indx2)>0 %both in the matrix
year_final(counter,1)=y;
quest_1F(counter,:)=quest_1(indx1,2:c1);
quest_2F(counter,:)=quest_2(indx2,2:c2);
e=ii;
if (counter==1)
s=ii;
end
counter=counter+1;
end
end
[r,c]=size(quest_1F);
K=s+2;
for i=1:r %each year
%Q1:
index_yes_1{i}=find(quest_1F(i,:)>=CAT); %students who answered yes to Q1
index_no_1{i}=find(quest_1F<CAT); %students who answered no to Q1
index_miss_1{i}=find(quest_1F(i,:)==0); %students who didn't answer Q1

%Q2:
if q2_RC==1
index_no_2{i}=find(quest_2F(i,:)==1);
index_yes_2{i}=find(quest_2F(i,:)==0);
index_miss_2{i}=find(quest_2F(i,:)==9);
else
index_yes_2{i}=find(quest_2F(i,:)==1);
index_no_2{i}=find(quest_2F(i,:)==0);
index_miss_2{i}=find(quest_2F(i,:)==9);
end

index_yes_both{i}=intersect(index_yes_1{i}, index_yes_2{i}); %students who said yes to both Qs
index_no_both{i}=intersect(index_no_1{i}, index_no_2{i}); %student who said no to both Qs
index_yes1_no2{i}=intersect(index_yes_1{i}, index_no_2{i}); %students who said yes to Q1 and no to Q2
index_no1_yes2{i}=intersect(index_no_1{i}, index_yes_2{i}); %students who said no to Q1 and yes to Q2
index_miss_both{i}=intersect(index_miss_1{i}, index_miss_2{i}); %students who left out both qs


total_yes_both(i)=length(index_yes_both{i}); %total who said yes to both questions
total_no_both(i)=length(index_no_both{i}); %total who said no to both questions
total_yes1_no2(i)=length(index_yes1_no2{i}); %total who said yes to Q1 and no to Q2
total_no1_yes2(i)=length(index_no1_yes2{i}); %total who said no to Q1 and yes to Q2
% Q1
%
% | yes | no
% ______|_______|_________
% Q2 yes | a | b
% ______|_______|_________
% no | c | d

a=total_yes_both(i);
d=total_no_both(i);
c=total_yes1_no2(i);
b=total_no1_yes2(i);

%formula for odds ratio:
OR= (a*d)/(b*c);

%find confidence interval for OR
lnOR=log(OR);
seOR=sqrt((1/a)+(1/b)+(1/c)+(1/d));
CI_upper=lnOR+1.96*seOR;
CI_lower=lnOR-1.96*seOR;
CI_upper=exp(CI_upper);
CI_lower=exp(CI_lower);
%odds_ratio cell matrix:
odds_ratio_cell{k,1}=[filename1];
odds_ratio_cell{k,2}=[filename2];
x=num2cell(OR);
odds_ratio_cell(k,K)=x;
OR_CI{k,1}=filename1;
OR_CI{k,2}=filename2;
upper=sprintf('%0.2f',round(CI_upper*100)/100);
lower=sprintf('%0.2f',round(CI_lower*100)/100);
OR_CI{k,K}=[lower ', ' upper];
K=K+1 ;
end
k=k+1;
end
end
cd ..
cd ..
cd OR_results
matsave=[ques1 '_response_' CAT '_OR_2013_cat.mat'];
matsave_CI=[ques1 '_response_' CAT '_OR_2013_cat_CI.mat'];
save(matsave, 'odds_ratio_cell');
save(matsave_CI, 'OR_CI');
cd ..
end

end
169 changes: 169 additions & 0 deletions odds_ratio/create_RR_OR_2013_CI.m
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,169 @@
%has reverse code information incorporated into this analysis

% This program goes through all the binary_NaN files and gets the relative risk of every combination
% of variables and displays them in a chart as follows:

% V1 | V2 | RR2001 | RR 2003 | RR2005 | RR2007 | RR2009 | RR2011
%____|____|________|_________|________|________|________|________
% | | | | | | |

% It then does the same thing for odds ratio


k=1; %counter for rows in rel_risk_cell and odds_ratio_cell

files1=dir(fullfile('C:','Users','rugglk01','Dropbox (Personal)','CDC','data','results_091614','NaN', '*.txt'));
cd ..
cd ..
cd matrices
load reverse_code.mat
cd ..
cd data
cd results_091614
cd NaN

P=length(files1);
for p=1:P
quest_1=importdata(files1(p).name, '\t');
filename1='';
a1=char(files1(p).name);
b1=strfind(a1,'-');
for q=1:(b1(1)-1)
c1=a1(q);
filename1=[filename1 c1];
end
ct=[];
for q=2:(b1(1)-1)
c1=a1(q) ;
if c1~=0
ct=[ct c1];
end
end
ct=str2num(ct);
q1_RC=reverse_code(ct,1);

%filename2
files2=dir(fullfile('C:','Users','rugglk01','Dropbox (Personal)','CDC','data','results_091614','NaN', '*.txt'));
N=length(files2);
for n=1:N
quest_2=importdata(files2(n).name, '\t');
filename2='';
a2=char(files2(n).name);
b2=strfind(a2,'-');
for m=1:(b2(1)-1)
c2=a2(m);
filename2=[filename2 c2];
end
ct=[];
for m=2:(b2(1)-1)
c2=a2(m);
if c2~=0
ct=[ct c2] ;
end
end
ct=str2num(ct);
q2_RC=reverse_code(ct,1);

%PROCESS THE COMBO
if ( isnan(q2_RC)==0 && isnan(q1_RC)==0) % won't compare a question to itself

[r1,c1]=size(quest_1);
[r2,c2]=size(quest_2);
year=[2001 2003 2005 2007 2009 2011 2013];
year1=quest_1(:,1);
year2=quest_2(:,1);
year_final=double.empty;
quest_1F=double.empty;
quest_2F=double.empty;
counter=1;
s=0; %start for 1
e=1; %start for 2
for ii=1:numel(year)
y=year(ii);
indx1=find(year1==y);
indx2=find(year2==y);
if numel(indx1)>0 && numel(indx2)>0 %both in the matrix
year_final(counter,1)=y;
quest_1F(counter,:)=quest_1(indx1,2:c1);
quest_2F(counter,:)=quest_2(indx2,2:c2);
e=ii;
if (counter==1)
s=ii;
end
counter=counter+1;
end
end
[r,c]=size(quest_1F);
K=s+2;
for i=1:r
%Q1:
if q1_RC==1
index_no_1{i}=find(quest_1F(i,:)==1); %students who answered yes to Q1
index_yes_1{i}=find(quest_1F(i,:)==0); %students who answered no to Q1
index_miss_1{i}=find(quest_1F(i,:)==9); %students who didn't answer Q1
else
index_yes_1{i}=find(quest_1F(i,:)==1); %students who answered yes to Q1
index_no_1{i}=find(quest_1F(i,:)==0); %students who answered no to Q1
index_miss_1{i}=find(quest_1F(i,:)==9); %students who didn't answer Q1
end
%Q2:
if q2_RC==1
index_no_2{i}=find(quest_2F(i,:)==1);
index_yes_2{i}=find(quest_2F(i,:)==0);
index_miss_2{i}=find(quest_2F(i,:)==9);
else
index_yes_2{i}=find(quest_2F(i,:)==1);
index_no_2{i}=find(quest_2F(i,:)==0);
index_miss_2{i}=find(quest_2F(i,:)==9);
end

index_yes_both{i}=intersect(index_yes_1{i}, index_yes_2{i}); %students who said yes to both Qs
index_no_both{i}=intersect(index_no_1{i}, index_no_2{i}); %student who said no to both Qs
index_yes1_no2{i}=intersect(index_yes_1{i}, index_no_2{i}); %students who said yes to Q1 and no to Q2
index_no1_yes2{i}=intersect(index_no_1{i}, index_yes_2{i}); %students who said no to Q1 and yes to Q2
index_miss_both{i}=intersect(index_miss_1{i}, index_miss_2{i}); %students who left out both qs


total_yes_both(i)=length(index_yes_both{i}); %total who said yes to both questions
total_no_both(i)=length(index_no_both{i}); %total who said no to both questions
total_yes1_no2(i)=length(index_yes1_no2{i}); %total who said yes to Q1 and no to Q2
total_no1_yes2(i)=length(index_no1_yes2{i}); %total who said no to Q1 and yes to Q2
% Q1
%
% | yes | no
% ______|_______|_________
% Q2 yes | a | b
% ______|_______|_________
% no | c | d

a=total_yes_both(i);
d=total_no_both(i);
c=total_yes1_no2(i);
b=total_no1_yes2(i);

%formula for odds ratio:
OR= (a*d)/(b*c);

%find confidence interval for OR
lnOR=log(OR);
seOR=sqrt((1/a)+(1/b)+(1/c)+(1/d));
CI_upper=lnOR+1.96*seOR;
CI_lower=lnOR-1.96*seOR;
%odds_ratio cell matrix:
odds_ratio_cell{k,1}=[filename1];
odds_ratio_cell{k,2}=[filename2];
x=num2cell(OR);
odds_ratio_cell(k,K)=x;
OR_CI{k,1}=[filename1];
OR_CI{k,2}=[filename2];
OR_CI{k,K}=[CI_lower ', ' CI_upper];

K=K+1 ;
end
k=k+1;
end
end
end

save('OR_2013', 'odds_ratio_cell');
save('OR_CI_2013', 'OR_CI');
Loading

0 comments on commit efd1762

Please sign in to comment.