using stset : no; data are mi set <- removing this setting if possible?

[Rose Matthews]

Error:

st survivalt, failure(fail==1)
no; data are mi set. Use mi stset to set or query these data ...
r119

(1)Hello I was trialling my code again on a new dataset I have just cleaned - which were patient outcomes data
(2) A few days ago as you have recognised from my posts I was asking about -multiple imputation - and was trying to learn on my stata.

I made a formal decision to first clean my patient outcomes data (1) - merge to my original data which I had been using to run this do file (2) - this never had - mi - stuff on here

I had been playing around with -mi- but never saved anything,
perhaps I may have -set- some settings from what I gather?

How do I remove this please? so I can just run st survivalt

The reason being, I haven't learnt multiple imputation fully and would like to apply once I have learnt it

I have no recollection which variable I used with miset or perhaps was trialling

[daniel klein]

Try

mi extract 0

[Rose Matthews]

hmm this doesn't work

I've also tried this

mi unset 

_mi_m not found 

mi set

data mi set mlong, M=0 
last mi update 16oct2023 15:44:53, approximately 6 hours ago

[Rose Matthews]

Hello, after lots of trying, I managed to solve it - but I still have one question - see bold below

I tried the following:

mi extract 0
no;dataset in memory has changes since last saved

mi unset
_mi_m not found

mi set

char list

Output from char list:
data[_mi_marker]: _mi_ds_1

Which apparently means Stata still thinks it's -MI- set

Then I tried:

char _dta[_mi_marker]

stset survival, failure(xx)
//this worked

Now, I don't know if I need to run char _dta[_mi_marker] everytime I run my do file ? Can anyone help in this regard?

I' haven't performed any imputations as seen in screen shot

[daniel klein]

Characteristics (char) get stored with the dataset. If you remove all _mi* characteristics before you save the dataset, you do not need to remove them again.