Thank You Prof! Do you have any reference that support this method?
@NirmalSingh-yy4lc
2 ай бұрын
ar5iv.org/pdf/2203.11820
@SimplytheBest23
Жыл бұрын
I have transformed my data to natural logs…after analysis how do I now revert back to original values?
@NirmalSingh-yy4lc
Жыл бұрын
Take the exponential of the series.
@rezkierdian402
2 жыл бұрын
Thanks! But, do you have any reference that support this method?
@elgrandosmokio1247
2 ай бұрын
Cari aja buku & artikel dar rick wicklin bro..
@01afmy
3 жыл бұрын
i have a question. if the data is in decimal point (example 0.0162), do we need to +1 or +0.1? tq
@NirmalSingh-yy4lc
3 жыл бұрын
I'm not sure, but I think that procedure will also not change the relative difference hence that can also be applied but in the situation you are talking the current procedure can be applied.
@AbdulhadisultaniA
2 жыл бұрын
Thank you! can we take the log of 13.94 and deducted it from the calculated log?
@NirmalSingh-yy4lc
2 жыл бұрын
In that case you can take log directly no need to adopt any additional procedure.
@inesgarcia3726
Жыл бұрын
hey thanks a lot! from where did you get this method?
@inesgarcia3726
Жыл бұрын
just as an update on this, I've just did it with eViews coding lw=@recode(w>0,log(1+w),-log(1-w)) and I've got the same result. so yes, this method works ! thank you again!!
@selmanaydemir
Жыл бұрын
thank you
@firstmotivation7564
3 жыл бұрын
Great
@philipelewis1
Жыл бұрын
This is truly nonsense. There is no such thing as a log of a negative number. The reason you take logs in the first place is calculate the fractional differences. Adding a constant to all the values to changes that relationship. Your procedure makes no sense whatever.
@NirmalSingh-yy4lc
2 ай бұрын
Thank you for your comment. I understand your concerns about taking the logarithm of a negative number, which is indeed mathematically undefined. However, in practice, a common method to address this issue is to add a constant to all values to ensure they are positive before taking the log. This technique, known as log transformation with a shift, is often used in research to stabilize variance and normalize data. It helps in better handling skewed distributions and can be particularly useful in time series analysis and other statistical methods. While it changes the absolute values, it preserves the relative differences, which is often what matters most in analysis. I hope this clarifies the rationale behind the procedure.
Пікірлер: 16