nielshintzen / vmstools

Automatically exported from code.google.com/p/vmstools
18 stars 12 forks source link

Linking logbooks and vms #10

Closed csilva53 closed 3 years ago

csilva53 commented 8 years ago

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

nielshintzen commented 8 years ago

Hi Cristina,

The merging function takes indeed FT_REF and the start-end time of the trip to link VMS, not the catch date. One can later on choose to use the splitAmongPings function to distribute the catches over the pings. This function can use a reference column, such as the time each VMS ping represents (pings with higher time associated will get a larger share). Similarly, if you have points at which it is more likely to catch fish, then make sure this indicator-column is used with this information per ping in it.

Hope this helps!

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 11:30 To: nielshintzen/vmstools Subject: [nielshintzen/vmstools] Linking logbooks and vms (#10)

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJIUIfrtEhS0nexwJp4OUU5PE05Hmks5qeZoEgaJpZM4Jg89y.

csilva53 commented 8 years ago

Hi Niels,

Thanks for your reply. Unfortunately, it doesn't solve my problem. One trip (FT_REF) may have more than 1 day and each day may have more than 1 haul (LE_ID). The lower levels that the function splitAmonPings consider are "trip", "day" or "ICESrectangle". In this case, I would like to split the haul catches by the time associated to the haul pings.

A solution might be to reconstruct the FT_REF to be an identifier of trip and haul, and then use the same procedure as with trip.

Regards,

Cristina

Às 12:44 de 10/08/2016, nielshintzen escreveu:

Hi Cristina,

The merging function takes indeed FT_REF and the start-end time of the trip to link VMS, not the catch date. One can later on choose to use the splitAmongPings function to distribute the catches over the pings. This function can use a reference column, such as the time each VMS ping represents (pings with higher time associated will get a larger share). Similarly, if you have points at which it is more likely to catch fish, then make sure this indicator-column is used with this information per ping in it.

Hope this helps!

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 11:30 To: nielshintzen/vmstools Subject: [nielshintzen/vmstools] Linking logbooks and vms (#10)

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJIUIfrtEhS0nexwJp4OUU5PE05Hmks5qeZoEgaJpZM4Jg89y.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/nielshintzen/vmstools/issues/10#issuecomment-238842655, or mute the thread https://github.com/notifications/unsubscribe-auth/AT-aUQ4m6v4vcCxQdQSfnz1LyGO2H9yLks5qebmTgaJpZM4Jg89y.

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

csilva53 commented 8 years ago

Again,

The pointed solution has some cons. The SI_DATIM is based on FT_DDAT and FT_DTIME. For the haul approach it should be based on the haul starting time. So it is not so straightforward as I initially thought.

Cristina

Às 14:23 de 10/08/2016, Cristina Silva escreveu:

Hi Niels,

Thanks for your reply. Unfortunately, it doesn't solve my problem. One trip (FT_REF) may have more than 1 day and each day may have more than 1 haul (LE_ID). The lower levels that the function splitAmonPings consider are "trip", "day" or "ICESrectangle". In this case, I would like to split the haul catches by the time associated to the haul pings.

A solution might be to reconstruct the FT_REF to be an identifier of trip and haul, and then use the same procedure as with trip.

Regards,

Cristina

Às 12:44 de 10/08/2016, nielshintzen escreveu:

Hi Cristina,

The merging function takes indeed FT_REF and the start-end time of the trip to link VMS, not the catch date. One can later on choose to use the splitAmongPings function to distribute the catches over the pings. This function can use a reference column, such as the time each VMS ping represents (pings with higher time associated will get a larger share). Similarly, if you have points at which it is more likely to catch fish, then make sure this indicator-column is used with this information per ping in it.

Hope this helps!

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 11:30 To: nielshintzen/vmstools Subject: [nielshintzen/vmstools] Linking logbooks and vms (#10)

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJIUIfrtEhS0nexwJp4OUU5PE05Hmks5qeZoEgaJpZM4Jg89y.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/nielshintzen/vmstools/issues/10#issuecomment-238842655, or mute the thread https://github.com/notifications/unsubscribe-auth/AT-aUQ4m6v4vcCxQdQSfnz1LyGO2H9yLks5qebmTgaJpZM4Jg89y.

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

nielshintzen commented 8 years ago

Hi Cristina,

Just to get things clear: what you want is to link a haul to specific vms pings, right? So, as you have multiple hauls in a day, you cannot assign just one haul to vms pings based on the control column but rather all hauls during a day. Correct?

Perhaps the easiest solution is to simply hack the code and add a level ‘haul’ to it and then run it again. The format is relatively straight forward. If you manage to get it coded correctly it would be a good idea to merge the code into the VMStools code. I think that I agree with you that right now there is no easy way of doing this (there certainly are complex ways, but the coding suggestion is that likely the more efficient one).

Regards,

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 15:23 To: nielshintzen/vmstools Cc: Hintzen, Niels; Comment Subject: Re: [nielshintzen/vmstools] Linking logbooks and vms (#10)

Hi Niels,

Thanks for your reply. Unfortunately, it doesn't solve my problem. One trip (FT_REF) may have more than 1 day and each day may have more than 1 haul (LE_ID). The lower levels that the function splitAmonPings consider are "trip", "day" or "ICESrectangle". In this case, I would like to split the haul catches by the time associated to the haul pings.

A solution might be to reconstruct the FT_REF to be an identifier of trip and haul, and then use the same procedure as with trip.

Regards,

Cristina

Às 12:44 de 10/08/2016, nielshintzen escreveu:

Hi Cristina,

The merging function takes indeed FT_REF and the start-end time of the trip to link VMS, not the catch date. One can later on choose to use the splitAmongPings function to distribute the catches over the pings. This function can use a reference column, such as the time each VMS ping represents (pings with higher time associated will get a larger share). Similarly, if you have points at which it is more likely to catch fish, then make sure this indicator-column is used with this information per ping in it.

Hope this helps!

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 11:30 To: nielshintzen/vmstools Subject: [nielshintzen/vmstools] Linking logbooks and vms (#10)

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJIUIfrtEhS0nexwJp4OUU5PE05Hmks5qeZoEgaJpZM4Jg89y.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/nielshintzen/vmstools/issues/10#issuecomment-238842655, or mute the thread https://github.com/notifications/unsubscribe-auth/AT-aUQ4m6v4vcCxQdQSfnz1LyGO2H9yLks5qebmTgaJpZM4Jg89y.

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10#issuecomment-238864878, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJCccytmW_lVf8Xi1OcDI0jazYWfnks5qedDGgaJpZM4Jg89y.

csilva53 commented 8 years ago

Thank you for the suggestions. Regards,

Cristina

Às 18:50 de 10/08/2016, nielshintzen escreveu:

Hi Cristina,

Just to get things clear: what you want is to link a haul to specific vms pings, right? So, as you have multiple hauls in a day, you cannot assign just one haul to vms pings based on the control column but rather all hauls during a day. Correct?

Perhaps the easiest solution is to simply hack the code and add a level ‘haul’ to it and then run it again. The format is relatively straight forward. If you manage to get it coded correctly it would be a good idea to merge the code into the VMStools code. I think that I agree with you that right now there is no easy way of doing this (there certainly are complex ways, but the coding suggestion is that likely the more efficient one).

Regards,

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 15:23 To: nielshintzen/vmstools Cc: Hintzen, Niels; Comment Subject: Re: [nielshintzen/vmstools] Linking logbooks and vms (#10)

Hi Niels,

Thanks for your reply. Unfortunately, it doesn't solve my problem. One trip (FT_REF) may have more than 1 day and each day may have more than 1 haul (LE_ID). The lower levels that the function splitAmonPings consider are "trip", "day" or "ICESrectangle". In this case, I would like to split the haul catches by the time associated to the haul pings.

A solution might be to reconstruct the FT_REF to be an identifier of trip and haul, and then use the same procedure as with trip.

Regards,

Cristina

Às 12:44 de 10/08/2016, nielshintzen escreveu:

Hi Cristina,

The merging function takes indeed FT_REF and the start-end time of the trip to link VMS, not the catch date. One can later on choose to use the splitAmongPings function to distribute the catches over the pings. This function can use a reference column, such as the time each VMS ping represents (pings with higher time associated will get a larger share). Similarly, if you have points at which it is more likely to catch fish, then make sure this indicator-column is used with this information per ping in it.

Hope this helps!

Niels

From: csilva53 [mailto:notifications@github.com] Sent: woensdag 10 augustus 2016 11:30 To: nielshintzen/vmstools Subject: [nielshintzen/vmstools] Linking logbooks and vms (#10)

I have checked the code of mergeEflalo2Tacsat and, as far as I undertood, the coupling is performed using the fishing trip reference (FT_REF) and date (LE_CDAT). I assume that the catches are distributed by the VMS pings according to the time assigned to each ping. Have you consider the allocation of catches by haul instead of date of catch? This is important when some species are mainly caught e.g. during the night or in a specific depth range.

Cristina

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10, or mute the

threadhttps://github.com/notifications/unsubscribe-auth/AEwaJIUIfrtEhS0nexwJp4OUU5PE05Hmks5qeZoEgaJpZM4Jg89y.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub

https://github.com/nielshintzen/vmstools/issues/10#issuecomment-238842655, or mute the thread

https://github.com/notifications/unsubscribe-auth/AT-aUQ4m6v4vcCxQdQSfnz1LyGO2H9yLks5qebmTgaJpZM4Jg89y.

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/nielshintzen/vmstools/issues/10#issuecomment-238864878, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AEwaJCccytmW_lVf8Xi1OcDI0jazYWfnks5qedDGgaJpZM4Jg89y.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/nielshintzen/vmstools/issues/10#issuecomment-238947178, or mute the thread https://github.com/notifications/unsubscribe-auth/AT-aUWsZ5D6PTSynKIR2MtmvfQp6wOGaks5qeg9IgaJpZM4Jg89y.

Cristina Silva Divisão de Modelação e Gestão de Recursos Pesqueiros Avenida de Brasília 1449-006 Lisboa @: csilva@ipma.pt mailto:csilva@ipma.pt

: +351 213027096 phoneto:+351213027096

nielshintzen commented 3 years ago

Partially solved, will be taken up with new development on haul info processing