Timeline for What kind of curve (or model) should I fit to my percentage data?
Current License: CC BY-SA 4.0
11 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Jun 24, 2020 at 5:23 | comment | added | Dennis | How did you do your equation search? | |
Jul 23, 2019 at 18:39 | comment | added | Carl Witthoft | @JamesPhillips I will attempt to do so (Heaviside --> errorbars or equivalent) | |
Jul 23, 2019 at 16:53 | comment | added | James Phillips | @CarlWitthoft because the full data set has now been added to the question, there is no need for conjecture - is now possible to determine this directly. For comparative purposes, please do so. | |
Jul 23, 2019 at 13:11 | comment | added | Carl Witthoft | Just to reiterate: application of, e.g., a Heaviside function, may yield similar error values. | |
Jul 22, 2019 at 20:30 | comment | added | James Phillips | Please note that as the actual data has now been added to the post, I have updated this answer using the updated data. | |
Jul 22, 2019 at 20:28 | history | edited | James Phillips | CC BY-SA 4.0 |
clarify
|
Jul 22, 2019 at 20:17 | history | edited | James Phillips | CC BY-SA 4.0 |
Added plot usine OP-edited data
|
Jul 22, 2019 at 16:32 | comment | added | James Phillips | Meh - this is semi-manually extracted data, the original data is required. This is usually sufficient for equation searches, and of course not for final results - which is why I said to use my extract-o-fit parameter values as initial parameter estimates on the original data. | |
Jul 22, 2019 at 16:27 | comment | added | mkt | There seems to have been an error in your data extraction: you have a bunch of negative percentage values. Also, your maximum values are at about 90% instead of 100% as in the original plot. You may have everything offset by about 10% for some reason. | |
Jul 22, 2019 at 16:26 | history | edited | James Phillips | CC BY-SA 4.0 |
clarify
|
Jul 22, 2019 at 16:20 | history | answered | James Phillips | CC BY-SA 4.0 |