Flat response curve problem

[Pritam chhetri]

Hello all,

I downloaded and clipped the bioclimatic data and some topographic data. Using my occurrence data and Bias file I ran the MaxEnt. I am getting all flat response curve for all the variables. What might be wrong that I am doing.  I have attached the screenshot my response curves. 

Thank you  

[Bede-Fazekas Ákos]

Hello Pritam,

It looks like your model thinks that every value (within the learnt interval) of all your predictors provide suitable environment for your species. Maybe the number of background points is very limited if compared to the number of presence points. You should check wether the study extent and the number of background points are large enough.

Have a nice week,
Ákos
----
Ákos Bede-Fazekas
Centre for Ecological Research, Hungary

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/3f019ba4-0b62-4e05-9d3f-f41e94ab41bfn%40googlegroups.com.

[Pritam chhetri]

Hello  Ákos,

Thank you for your response. I have used 10000 background points with a 30% random seed. The feature types used were linear and quadratic. I have only 35 occurrences of which 10 occurrences were used in testing the model and 25 were used in training the model. I am using 30 seconds resolution raster data. I have attached the HTML file that MaxEnt generates. Please have a look at what is wrong with my data or model. 

Thanks and Regards,

Pritam

[Bede-Fazekas Ákos]

Hello Pritam,
In this case I have no other idea, sorry.
Have a nice weekend,
Ákos

To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/a3695947-0e25-435b-ab7c-7bdda18b8e9dn%40googlegroups.com.

[Amir Sohail Choudhury]

Hello,

Try 'leave one out' (n-1) method for Cross validation. 

I think, it will solve, try once. 

Best regards

Amir

To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/a3695947-0e25-435b-ab7c-7bdda18b8e9dn%40googlegroups.com.

[Pritam chhetri]

Hi Amir,

I tried that method too but the response curve is still the same.

Thank you

[Sunil Sapkota]

Dear pritam, 

For me there is error while you resample the data. I noticed similar error in my practice. 

Have a great day

--

[Pritam chhetri]

Hi Sunil,

Thank you for your response. I am not clear on what resample you are talking about. Please explain to me in detail.

Many thanks

Pritam

[arjun nepal]

Hi Pritam

While resampling sometimes you will notice that all the data present in ASCII is 999. If you have only 999 in your dataset it may be due to conversion process ( While changing the extent and resolution of image). 

Thanks and Regards

To view this discussion on the web visit https://groups.google.com/d/msgid/maxent/c78588df-7006-487a-bbc2-bf1c99f58657n%40googlegroups.com.

[Pritam chhetri]

Hi Arjun,

I checked the values of the ASCI files and it has a different values. Please have a look at the attached screenshot below.

Thank you