24-03-2012, 09:34 AM
Hi Pansy Mae,
The BreastNexus e-mail facility doesn't support uploading files either, but step by step, we're getting there.
When I plot the difference between bust and underbust versus time, there is scatter. I can reduce the scatter if I lower the breast sizes measured at high BMI, and increase the breast sizes measured at low BMI. This "correction" is no higher than 1 cm per BMI point. I don't know if my excel starter version on this netbook supports a linear regression of
Breast-Underbust = a*time + b*BMI + c
but I'll try it later.
I admire your discipline in achieving the weight loss. However, it's quite rapid. If you would apply the above correction to your data, you will probably notice that a, the underlying real growth, is significantly higher than zero.
In my rapid weight loss during the past month, I noticed firmer breasts too. The problem in your data set is that time and BMI are no independent variables, as you have consistently lost weight over time. Once we get the math right, and estimate b, you will find a.
The BreastNexus e-mail facility doesn't support uploading files either, but step by step, we're getting there.
When I plot the difference between bust and underbust versus time, there is scatter. I can reduce the scatter if I lower the breast sizes measured at high BMI, and increase the breast sizes measured at low BMI. This "correction" is no higher than 1 cm per BMI point. I don't know if my excel starter version on this netbook supports a linear regression of
Breast-Underbust = a*time + b*BMI + c
but I'll try it later.
I admire your discipline in achieving the weight loss. However, it's quite rapid. If you would apply the above correction to your data, you will probably notice that a, the underlying real growth, is significantly higher than zero.
In my rapid weight loss during the past month, I noticed firmer breasts too. The problem in your data set is that time and BMI are no independent variables, as you have consistently lost weight over time. Once we get the math right, and estimate b, you will find a.