I am new to stata and my work is about medical data.
My data are about diagnostic performance of different imaging modalities (I am using random data in the example). I have installed a stata extension called diagnostic test (the command is “diagt”, first variable is reference standard second the test, both are binary). In my analysis diagt is executed many times and I want to save/store the r-class data/values provided by the test e.g. sensitivity and specificity in order to summarize them later (at the end of the “do” file) in a table. I guess that diagt is not important and that it would be the same for all commands providing r-class data.
At the moment I store the values in macros and put the macros at the end in a matrix which I display with “matlist”.
My questions:
Is there an easier/more elegant way to store the r-class values and show them in a table afterwards?
It seems not optimal that I must specify each value I want to use later and put it into a macro. Is it possible to store all the r-class data at once and display specific values later?
Code:
// run diagt diagt fasitb MRtb // save some r-class data global MRt_sens = r(sens) global MRt_spec = r(spec) // run diagt for another modality/test diagt fasitb MRnb // save the same r-class data into different macros global MRn_sens = r(sens) global MRn_spec = r(spec) // now I generate my matrix and the summary table matrix A = $MRn_sens, $MRn_spec\ $MRt_sens, $MRt_spec \ $MCC_overall, . \ $MCC_sens, $MCC_spec matrix rownames A = "MR nativ" "MR toilax" "McNemar Overall, Nativ vs Toilax" "McNemar sens spec" matrix colnames A = sensitivity specificity matlist A, format(%15.1f) twidth(40) title(Summary of Results)
Array
Here a list over all r-class data provided by diagt and I will probaly use several of them.
Code:
return list scalars: r(sens) = 50.12003200853561 r(sens_lb) = 48.50661567935583 r(sens_ub) = 51.73326103478671 r(spec) = 49.96000639897616 r(spec_lb) = 48.71277950932287 r(spec_ub) = 51.20727063140053 r(prev) = 37.49 r(prev_lb) = 36.53994481639038 r(prev_ub) = 38.44734619397738 r(ppv) = 37.52746155382465 r(ppv_lb) = 36.18391913702044 r(ppv_ub) = 38.8855612589619 r(npv) = 62.54756659323052 r(npv_lb) = 61.1880343889589 r(npv_ub) = 63.89241263219665 r(lrp) = 1.001599488763926 r(lrp_lb) = .9619277032697625 r(lrp_ub) = 1.042907416515918 r(lrn) = .9983979504151261 r(lrn_lb) = .9587131869858447 r(lrn_ub) = 1.039725416239468 r(or) = 1.003206675602117 r(or_lb) = .9251808504343817 r(or_ub) = 1.087812866382799 r(roc) = .500400185585022 r(roc_lb) = .490277306910667 r(roc_ub) = .510523064259377 r(oprev) = . r(oprev_lb) = . r(oprev_ub) = . r(opos) = . r(opos_lb) = . r(opos_ub) = . r(oneg) = . r(oneg_lb) = . r(oneg_ub) = .
0 Response to Store r-class values and tabulate later
Post a Comment