Changing a large number of absorption coefficients

Comments

27 comments

  • Avatar
    DougO
    You can do it easily with MS Access. First build your model through the Social Accounts. Close your model and then you can open your model file directly in Access. The absorbption coefficients are in the Regional Absorption table. Make your edit changes, you can use a query to get just the industries you want. Or copy and past the whole thing into Excel. Or create a query to update the Regional Absorption table. Once the changes are made, close that table and open the Model Specs table. Make an entry in the Detail field called Dirty and set the Value to 2. Open your model up and rerun the social accounts and it will start after the production function change.
    0
    Comment actions Permalink
  • Avatar
    DougO
    You can do it easily with MS Access. First build your model through the Social Accounts. Close your model and then you can open your model file directly in Access. The absorbption coefficients are in the Regional Absorption table. Make your edit changes, you can use a query to get just the industries you want. Or copy and past the whole thing into Excel. Or create a query to update the Regional Absorption table. Once the changes are made, close that table and open the Model Specs table. Make an entry in the Detail field called Dirty and set the Value to 2. Open your model up and rerun the social accounts and it will start after the production function change.
    0
    Comment actions Permalink
  • Avatar
    Unknown User
    :( Either I or one of my colleagues has mistakenly adjusted the US Absorption Coefficient table instead of the Regional Table in MS Access. Although the underlying odf files are not adjusted, whenever a new model is built the production function is already adjusted. How do we restore the default production function?
    0
    Comment actions Permalink
  • Avatar
    Unknown User
    :( Either I or one of my colleagues has mistakenly adjusted the US Absorption Coefficient table instead of the Regional Table in MS Access. Although the underlying odf files are not adjusted, whenever a new model is built the production function is already adjusted. How do we restore the default production function?
    0
    Comment actions Permalink
  • Avatar
    DougO
    The us absorption and byproducts tables are linked rather than imported into the model. So editing the model's US absorption and by-products edits the default structural matrices. Re-download and re-install the 2006 structural matrices (06nat509.ims).
    0
    Comment actions Permalink
  • Avatar
    DougO
    The us absorption and byproducts tables are linked rather than imported into the model. So editing the model's US absorption and by-products edits the default structural matrices. Re-download and re-install the 2006 structural matrices (06nat509.ims).
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Doug, If there is possible to do this on V3 implan? I have the same issue, a large number of absorption coefficient for some industry to be changed. I have tried to do this with the V3. It imports successfully. I saw the new absorption coefficients. However, when I run again the model, the absorption coefficients change to the original ones. I think there is something related with the "click" FIXED option that is for every coefficient changed. Therefore, I have two questions: 1) if there is possible to keep doing this? 2) if yes, if the problem is the "FIXED" option, if there is any way I can upload the modified file and that the "fixed" option appear clicked? I manually click on all coefficients in the IMPLAN model and it works. I just want to find a way to do it faster. Thanks for any help provided.
    0
    Comment actions Permalink
  • Avatar
    jenny
    Could you please describe exactly what steps you are taking and what you are trying to accomplish? For instance, what exactly are you importing? Are you working from within the software or in the Access database? Thank you.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Yes. Thanks. I have interest in solving problems A and B. However, my urgency now is on item A. Any help will be appreciate it. A) I want to change the absorption coefficients for 5 industries. For that I did the following: 1) create my model in IMPLAN. I am using county level data.I also have to click on every tab under Explore>Social Accounts and Explore>Industry Accounts. Otherwise, I will not be able to see the Regional Absorption tables in the Access File. 2)close the model 3)open the model in access 4) modify the column gross absorption coefficients under RegionalAbsorption Table 5) save it 6) open the program again on IMPLAN. What is happening is that the program upload the new coefficients, but once I construct the model again (to have SAM with new values), the program provided me with the original absorption coefficients again. So, what I have to do, which is a tedious process, has been: 7) for every sector that I have changed the coefficients (in my case 6 industries), import the production function. 8) construct the model 9) for EVERY Industry that I want to change the coefficients, export the production function that I had imported under item 6) and construct the model 9a) I found that 7--> 9 is a tedious process. I have not been able to fix it with anything else. Following Doug comments, I have also changed the column value to 2 under the "Dirty" column on the table "Common Model Specifications" on Access. However, once I construct the model again, I have the original absorption coefficients again. So, my following questions are: 1) if there is any way to create the regional absorption, valued added and byproducts tables without going through step 1 in order to see these tables in Access? 2) if there is an easier way to change the absoprtion coefficients? what I meant by that is avoid steps 7-->9. B) The other thing I would like to explore is to change absorption AND valued added coefficients. Therefore, what is the best way? I know that the total gross absoprtion coefficients and the valued added have to add up to 1. What I want to do then, it to change both at the same time. 1) I have changed the coefficients under regional absoption and regional valued added coefficients on the respectively tables on Access (RegionalAbsorption and RegionalValuedAddedCoefficients). For it, I follow the same steps under items 1-6 2) I have tried to run it with a without item 9a. It did not work. My model does not upload the regional added coefficients. So I was only able to change the regional absorption coefficients (following items 7-->9). If I want to change the valued added values I have to do it by hand under the IMPLAN tab Customize--> Study Area Data for each industry (in my case 6).Besides the fact that that is a tedious process, the problem with this is that my regional absoprtion coefficients change to the original values once I construct the model again.
    0
    Comment actions Permalink
  • Avatar
    DougO
    The way the flags are set in the software, this is the order you will have to do things to not lose your Access edits: 1) Create a new model (don't create social accounts or multipliers) 2) Load the value added, output, employment changes in the access model 3) Create regional absorption coeficients through the software (Options > Construct > Social Accounts > Regional Absorption 4) Customize regional absorption coeficients through the software (this is a dummy edit to set flags, so only change a single item) Choose an industry whose coefficients you are going to replace anyway. Balance and save the edit. 5) Load your absorption coefficients into the access model. 6) Create the multipliers. The flags set in step 4 will not allow the software to overwrite the changes you load into the data base.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Thanks Doug for the answer. I was not able to follow it. On item 2, what you meant is to upload my new coeffs into the RegionalAbsorption and RegionalValuedAddedCoefficients Tables in Access, right?.Given that I did not construct the SAM, these tables were originally empty. I did item 3, however the original absorption coefficients appear again. I dont know if I follow your instructions from item 4. What I did was to choose one sector that I change the absorption coefficients and just change one number ONLY (e.g. 3001 commodity from 0.066801 to 0.066802). Then I clicked on balance and then save it. Here the absoption coefficients still were the original ones. Given that the absorption coefficients that appear are the original ones, I am not able to upload mine on item 5. The absorption coefficients I built include new commodities for some industries I am modifying. As I have the original coefficients on item 6, my new aggregated SAM (which is my interest) is the same as the original. Thanks Thanks
    0
    Comment actions Permalink
  • Avatar
    DougO
    Since you are changing production function I assumed the industry components - value added, employment, etc - were being edited as well (as was done in the beginning of this forum thread). The StudyArea tables hold this data and are created as soon as the model is created. If not then skip step two - although you will need to make sure that your absorption coeficients plus the value added coeficients sum to 1.0, otherwise the industry sector will not be balanced.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Thanks Doug. My interest was on both valued added and absoprtion coeffs. I was doing the changes of valued added under the RegionalValuedAddedCoeffs but not under the Study Area tables. I guess there is no way to avoid the procedure on item No 4. I was thinking to link Access with IMPLAN in VBA or so in order to make a program for a user that does not have a lot of knowledge of neither IMPLAN or ACCESS. I dont think that is possible,, does it? Thanks Doug
    0
    Comment actions Permalink
  • Avatar
    DougO
    There is one slight alteration that will work, but don't think it is any easier. In step 3, run not only "Regional Absorption", but also "Regional Byproducts" & "Commodity Accounts" (but not "Social Accounts(SAM)"). Then you may skip 4 and go to step 5 which is to enter the absorption coeficients into Access. Step 6 will only run the "Social Accounts (SAM)" and "Multipliers" routines in the software which will not reload the coeficients.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Good morning. I want to return to this issue. I followed the instructions. Specifically, what I did was: 1. Create a new model into IMPLAN. Don’t click on Construct Social Accounts or Multipliers. 2. Open the IMPLAN file in ACCESS Office. 3. Load the value added values into the ACCESS table “StudyAreaValueAdded”. 4. Create regional absorption coefficients, regional byproducts and commodity accounts through the software (Options > Construct > Social Accounts > Regional Absorption, Regional Byproducts, and Commodity Accounts) 5. Load the absorption coefficients into the ACCESS table “RegionalAbsorption” 6. Construct the social account, the industry account and the multipliers through the software (Options > Construct > ) As you can see, the only changes I had made are the gross absorption and the valued added. I made sure that the gross absorption and the valued added coeffs add up to 1. However, when I check the aggregate SAM I found that the model is not balanced. I have attached the excel file. One sheet contains the original SAM without any changes "Regional SAM Balances Original". The second sheet contains the SAM modified by the changes in gross absorption and valued added. The line 44 in the second sheet shows the difference between the sum of the columns and the row. These differences are attributed to the changes I have made. What could be the reason of this difference? Any support will be really appreciated it. Thanks in advance.
    0
    Comment actions Permalink
  • Avatar
    DougO
    When we create the study area data and run the gravity model for version 3, the domestic trade flows fit with the data so so that the model is balanced. When you make changes to value added and no corresponding changes to Final demand the model becomes unbalanced as domestic trade flows are fixed in version 3. If you were to rerun the model, just the trade flows in social accounts, using econometrics, I believe the model would balance itself as domestic trade flows would be forced to balance the model. Second thoughts - only imports are fixed but balancing may not work even with econometrics as domestic exports would not be allowed to go negative depending on the changes. I assume that the worksheet is an aggregation of your data. In the detailed version are domestic exports 0 for unbalanced sectors?
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Thanks Doug. I did not follow this phrase "I assume that the worksheet is an aggregation of your data. In the detailed version are domestic exports 0 for unbalanced sectors?". Can you clarify it for me? However, I have made some modifications in the valued added values (I had modified 7001 instead of 6001 basically), re-run and it seems to work better. I have to check again what I did differently this time in relation to the previous time. Now I want to complicate things a little bit more. Therefore some more questions. 1) Can I alter the RPCs in the gross absorption matrix table in ACCESS? is the table “RegionalAbsorption” the best way to do it? if it is, once I run IMPLAN (with the gravity model)are my modified RPCs will be limited by the RSC as IMPLAN usually does? 2) Is the table "RegionalInstitutionDemand" the best table to introduce changes in final demand to households using ACCESS? Or do I have to use "StudyAreaFinalDemand". I have found that IMPLAN is using households as a balance mechanism(specifically modifying the rows and columns for households to force balance). Is that always true? If that is the case, I am afraid that once my changes in final demand for households are incorporated, there will be lost once IMPLAN balance it. Just wondering. 3) If I want to modify RPCs of institutions (have different values for institutions and industries) can I do it in the "RegionalInstitutionDemand"? It is my guess that the same procedure applied in item (1) will work for this one as well. Finally, if there is a documentation that talks more about ways to induce large numbers of changes in different values and/or coefficients using ACCESS? Sometimes I don not know when to impose the changes in the "StudyArea" ACCESS tables or into "Regional" ACCESS tables. Thanks so much Doug.
    0
    Comment actions Permalink
  • Avatar
    DougO
    Are you working with 440 sectors (the SAM you sent had far fewer so it was aggregated) or did you aggregate the whole model and not just the report? If you look at commodity balances (explore > social accounts > commodity balances) do the detailed sectors for the sectors you changed have a domestic export of 0? The other cause of unbalanced SAM might be if the sum of the modified coeficients in the absorption matrix plus the sum of modified value added coeficients do not sum to 1.0000000 The procedure is similar as before but the addition of modifying trade flows changes things a bit: The way the flags are set in the software, this is the order you will have to do things to not lose your Access edits: 1) Create a new model (don't create social accounts or multipliers) 2) Load the value added, output, employment changes in the access model 3) Create regional absorption coeficients through the software (Options > Construct > Social Accounts > Regional Absorption 4) Customize regional absorption coeficients through the software (this is a dummy edit to set flags, so only change a single item) Choose an industry whose coefficients you are going to replace anyway. Balance and save the edit. 5) Load your absorption coefficients into the access model. 6) Construct the model through "Commodity Accounts" (Options > Construct > Social Accounts > Commodity Accounts 7) Do a dummy edit of trade flows to set the flags - only change a single item as before. 8) Create the multipliers. The flags set in step 4 and 6 will not allow the software to overwrite the changes you load into the data base. Note, when I tried this the tested RPC entered appeared to be constrained. You will need to modify StudyAreaFinalDemand - households are balanced through Capital ("savings") - if you net increase the spending you will decrease savings. Yes, RPCs for institutions would be modified at the same time you modify RegionalAbsorption as described above. No documentation, we do not encourage modification of the model through Access.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    To clarify. 1) It is in step 5 where I have to modify the RPCs for industries besides the gross absorption, right? 2) I thought that you were able to have different RPCS for institutions and industries. I would like to have the option to upload different RPCs for institutions (households) and industries in the ACCESS tables. Is that possible? Thanks Sara
    0
    Comment actions Permalink
  • Avatar
    DougO
    No, sorry, it would be step 7.5 Load in rpcs into RegionalAbsorption. After 7, the dummy trade flow edit. I forgot to include the step. The institutional rpcs can be modified as part of step 7.5 in table RegionalInstitutionDemand.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Ok. I think I am following you now. In one of your previous answers you wrote that I could avoid to do step 4 to "set the flags" The steps I have used are described on reply #12914. As you will see there was no need to do step 4. I think the procedure I implemented (based on your notes) worked fine. In the same line, can we avoid "to set the flags" in the case of trade flows or step 7?. Thanks
    0
    Comment actions Permalink
  • Avatar
    DougO
    If you skip step 4 and recreate multipliers, the software will not use your changes to create the multipliers. The "commodity balances" step is where the regional absorption data is integrated into the process of creating multipliers. However, running that step will re-grab the original absorption coefficients unless you also do step 4. You will have to follow the 8 (now 9) steps as outlined to keep and use the customized absorption and rpc data.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Doug, forgot to answer your question. "Are you working with 440 sectors (the SAM you sent had far fewer so it was aggregated) or did you aggregate the whole model and not just the report? If you look at commodity balances (explore > social accounts > commodity balances) do the detailed sectors for the sectors you changed have a domestic export of 0?" I wanted to follow your step (explore > social accounts > commodity balances). I am not able to find "commodity balances". Either way,is not exports estimated in terms of commodities? the gross absorption and valued modifications were at the industry level. One more time, I am missing something here. Thanks
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Doug. Sorry again for so many questions. One last point. I found a possible bug maybe due to the changes in parameters (e.g. gross absorption, valued added, etc). You have to correct me if I am wrong. But I think this bug does not affect the results. I have attached a word document that contains print screens of different tabs from IMPLAN. I am using industry 62 to show the templates. 1) If you go through CUSTOMIZE>STUDY AREA, the modified parameters are there. 2) If you go through EXPLORE>SOCIAL ACCOUNTS or EXPLORE>INDUSTRY ACCOUNTS and explore an industry with modified parameters, the modified valued added values are also there. 3) if you go through EXPLORE>STUDY AREA DATA, the parameters and values are the ORIGINAL ones. 4)The ACCESS table StudyAreaIndustryData has also the original values The fact that (3) and (4) has original values do not affect the final estimates, isnt it? I found that the multipliers resulted from the modified parameters are different from the original multipliers. So my guess is that IMPLAN used information from either (1) and (2) to obtain the multipliers and final AGGREGATED SAM tables. Am I wrong? Thanks so much
    0
    Comment actions Permalink
  • Avatar
    DougO
    Sorry, that was V2 terminology, the report name is "Commodity Summary". Another item to look at is the Explore > Social Accounts > Balance Sheet > Industry Balance Sheet, and choose an edited industry. Make sure that the Code 3000 Total Commodity Demand; Gross Absorption % (commodity demand tab) plus the Total Value Added: Value Added Coefficient % (Value Added tab) sum to 1.0. The Explore > study area data > Industry Detail report is pulled from table StudyAreaIndustryData, which the software does not know to update if you edit a study area through Access. That table is only used for display and does not enter into the process of creating multipliers.
    0
    Comment actions Permalink
  • Avatar
    salvalizarraga
    Thanks Doug. That was a relief. I have checked the industries I have altered the gross absorption and valued added in the IMPLAN program. They added up to 1. I have also checked that the sum of the ACCESS tables RegionalAbsoption + RegionalValuedAddedCoeffs add up to one. Since they showed the results from the modification of parameters, I thought it was a good idea. All the industries add up to 1 or 1.000000009 or 0.9999999999997. They also match the implan numbers shown in the tabs that you mentioned. The IMPLAN program does not show those decimal numbers tough. However, I think it is just a rounding issue. I guess the IMPLAN algorithm try to make sure these sum always add up to one or very very close to it.Together with all the other algorithms, IMPLAN try to make sure that the sum of the columns and the rows for each industry, commodity, institutions, and so on are equal or very close to it. I dont think I should worry for it. Correct me if I am wrong If I am correct I am satisfied with the procedure. Thanks Doug. I am going to try to work with the procedure you told me to modified the RPCs. I will get back to you if I find problems. Once again thanks.
    0
    Comment actions Permalink
  • Avatar
    DougO
    I think you are set. If you get a chance, send me the model and tell me which sectors you modified so I can check into why it doesn't balance. Thanks.
    0
    Comment actions Permalink

Please sign in to leave a comment.