Serious Bug with MRIO

Comments

9 comments

  • Avatar
    BECAnalyst
    I am available to do a LogMeIn session right now to show you the problem...
    0
    Comment actions Permalink
  • Avatar
    BECAnalyst
    Could it have something to do with the fact that I'm using the same state-level model to link in to each county-level model? It seems to always display the MRIO state-level results for the last scenario that I've run that links in that state-level model.
    0
    Comment actions Permalink
  • Avatar
    Frances
    Hi Michelle, Just to verify based on your last post, it sounds as if you are linking the same models together multiple times (i.e. you have one set of county and state models and you are interlinking them to each other, as opposed to having state model A linked to county models A, state model B linked to county models B, state model C linked to county model C). For a variety of reasons each set of MRIO linkages must occur in a distinct subset of models. So when region A is linked to B, C, and D. I new linkages are setup from B new models B1, A1, C1 and D1 need to be created to examine these impacts. If you are not create a discreet set of models for each group of linkages, this is likely the reason that you are encountering this issue. If you are creating discreet sets of models for each linkage, please let me know, and we can look into this further. Frances
    0
    Comment actions Permalink
  • Avatar
    Frances
    Also just to ensure that the MRIO's are correctly assembled... I presume the counties you are linking to state model are counties from regions outside of the state? If not (i.e. you are linking counties to the state model they occur in geographically speaking) this is double counting the MRIO impacts. For MRIO between all the counties in the state you would need to build a model including the county of interest and then MRIO models from the remaining counties representing the "rest of the state" less the original study area region. Frances
    0
    Comment actions Permalink
  • Avatar
    BECAnalyst
    I have counties A, B, C and D and state E. My scenarios take place in each of the county level models. I'm linking A to E, B to E, C to E and D to E. But you're saying that for each county of A, B, C, and D I need to have a separate state E?
    0
    Comment actions Permalink
  • Avatar
    Frances
    Exactly if E needs to be linked to each county model you would need a separate model E for each time you link it to a county (i.e. A-E1, B-E2, C-E3, D-E4) So if the county models aren't being used in other linkages you'll only need to build those once, but you will need for E models to avoid any issues. Frances
    0
    Comment actions Permalink
  • Avatar
    BECAnalyst
    Say I have county A and I want to run two scenarios, both of which involve linking in state B. Do I need separate state B models for each scenario as well?!
    0
    Comment actions Permalink
  • Avatar
    Frances
    No you shouldn't. You should be able to run as many impact scenarios on a set of linkages as you want as long as all the Scenarios are unidirectional A -> B its only when they direction changes B -> A that you would need to create two new models. Frances I should also add, and that the models are not attached to any other models. So both A and B would need to models used in no other linkages.
    0
    Comment actions Permalink
  • Avatar
    Frances
    Yep once a model is linked to another model they exist only for each other. They cannot be used for any other linkages. So for example say I need to link each county to E and then I need also to examine how A links to B-d and how B links to A, C, and D. I would need to create the following models/linkages. A1-E1, B1-E2, C1-E3, D1-E4 A2-B2, A3-C2, A4-D2 B3-A5, B4-C3, B5-D3 Typically we recommend creating these models in different folders to keep the linkages straight. But yes in the scenario listed above you would need 4 E models 5 A models
    0
    Comment actions Permalink

Please sign in to leave a comment.