Timeline for Is my logistic regression model correct?
Current License: CC BY-SA 4.0
19 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Oct 3 at 10:03 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
Jun 1 at 8:00 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
Jan 25 at 8:08 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
Sep 22, 2023 at 9:06 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
May 12, 2023 at 6:02 | history | bumped | CommunityBot | This question has answers that may be good or bad; the system has marked it active so that they can be reviewed. | |
Jul 28, 2020 at 3:00 | history | tweeted | twitter.com/StackStats/status/1287945901319888897 | ||
Jul 27, 2020 at 20:23 | answer | added | eithompson | timeline score: 0 | |
Jul 27, 2020 at 19:26 | comment | added | eithompson | @MustaphaHakkouAsz this doesn't answer your question, but for future reference you can use y ~ A*B as your formula: it is shorthand for y ~ A + B + A:B | |
Jul 27, 2020 at 19:24 | comment | added | eithompson | @Dave2e when an N-level factor is put into the glm() formula argument, it is automatically separated into N-1 binary variables for the regression (one-hot encoded). So even if OP has them coded as factors, then there should be no difference between the binary encoding. | |
Jul 27, 2020 at 17:56 | answer | added | dimitriy | timeline score: 0 | |
S Jul 27, 2020 at 16:22 | history | suggested | Nuclear241 | CC BY-SA 4.0 |
Minor edit to improve readability
|
Jul 27, 2020 at 14:45 | review | Suggested edits | |||
S Jul 27, 2020 at 16:22 | |||||
Jul 22, 2020 at 20:23 | comment | added | Dave2e | This is just a possible explanation and I have not verified it. If you coded A and B as factors than most likely R treated those values as a 1 and 2 and not your original 0 and 1. | |
Jul 22, 2020 at 17:57 | comment | added | Mustapha Hakkou Asz | yes they are : A takes high or low and the same for B its a 2*2 Design | |
Jul 22, 2020 at 17:54 | comment | added | Harvey Motulsky | To clarify: factor A, factor B, and Y all are binary with two levels? | |
Jul 22, 2020 at 17:47 | comment | added | Mustapha Hakkou Asz | Y is not continuous it is a response taking 2 values 1 or 0 | |
Jul 22, 2020 at 17:41 | review | First posts | |||
Jul 22, 2020 at 18:50 | |||||
Jul 22, 2020 at 17:38 | comment | added | Harvey Motulsky | Logistic regression is used when the outcome (Y) is binary. Is that the case for your data? I doesn't seem so. If Y is continuous, then logistic regression is the wrong tool. You may be confusing it with fitting a logistic model using nonlinear regression. | |
Jul 22, 2020 at 17:34 | history | asked | Mustapha Hakkou Asz | CC BY-SA 4.0 |