Successfully removed.
Sorry, an error occurred.
|
madlopez22 favorited a video
(1 day ago)
|
|
| |
|
|
madlopez22 uploaded a new video
(4 days ago)
|
|
| |
|
|
madlopez22 uploaded a new video
(1 week ago)
|
|
| |
|
|
madlopez22 favorited a video
(2 months ago)
Calculation of the General Correlation coefficient (invention). I also n...
more
Calculation of the General Correlation coefficient (invention). I also named it the Monotonic Correlation Coefficient. It uses the sum of products.
When the relationship is perfectly distributed around the strait line, which occurs extremely rarely, the monotonic corr. and r are equals. Usually, while close, they are not equal. All thousands tests (n=10 to 1000) done with linear and non-linear relationships, the monotonic r is very close to Pearson r if the relationship is linear but is higher than Pearson when the relationship is curvilinear (ex.: exponential or logarithmic relationship) because these type of curves are part of "monoticity" domain while they are outside the linearity domain per definition. Let it be clear, the Mono. corr is not the Pearson r. I do not know any other way to calculate the Mono. corr. other than the algorithm I developed (shown in this video). If you want to a SAS macro that calculates the Monotonic coefficient and compare it with other correlation coefficients, let me know (mletou@rael.org), I will send it to you.
less
|
|
| |
|
|
madlopez22 liked a video
(2 months ago)
Calculation of the General Correlation coefficient (invention). I also n...
more
Calculation of the General Correlation coefficient (invention). I also named it the Monotonic Correlation Coefficient. It uses the sum of products.
When the relationship is perfectly distributed around the strait line, which occurs extremely rarely, the monotonic corr. and r are equals. Usually, while close, they are not equal. All thousands tests (n=10 to 1000) done with linear and non-linear relationships, the monotonic r is very close to Pearson r if the relationship is linear but is higher than Pearson when the relationship is curvilinear (ex.: exponential or logarithmic relationship) because these type of curves are part of "monoticity" domain while they are outside the linearity domain per definition. Let it be clear, the Mono. corr is not the Pearson r. I do not know any other way to calculate the Mono. corr. other than the algorithm I developed (shown in this video). If you want to a SAS macro that calculates the Monotonic coefficient and compare it with other correlation coefficients, let me know (mletou@rael.org), I will send it to you.
less
|
|
madlopez22