Simple data manipulation problemthat I just cannot get past! I have a dataset with the following variables: A B C Iter 1 2 9 1 2 5 7 1 3 9 6 1 4 1 3 2 I also have another dataset with just one row of info. A_A B_B C_C 1 3 7 what i need to do is the following: For ITER = 1 First A_A_P1 = A_A + A -> 1 + 1 = 2 For subsequent ITER = 1 A_A_P1 = lag(A_A_P1) + A -> 2 + 2 = 4 So the A_A_P1 column should look like this: A_A_P1 2 4 7 11 For iter = 2, same process but now the values will be under A_A_P2. The same process will continue for B and C with respective ITERs. Any idea how this can be done efficiently?
... View more
Rookie here, I am trying to replicate the study in the following paper: http://support.sas.com/resources/papers/proceedings16/SAS6364-2016.pdf The process I followed is: -Run Proc Varmax to model BVAR on log-ratio of (RealGDP, CPI and Unemployment Rate). Save residuals in a separate dataset. -Call Proc Copula to find he joint distribution from the marginal distribution of the residuals. -Once I get that, I use the simulated errors with the forecast means from Varmax to create macroeconomic scenarios. I have the following issue. My residuals do not follow normal dist. I am not sure how to transform them. Proc Copula does not seem to work with the assumptions. Any idea how I can fix this? Your help is much appreciated I am using the following code. PROC COPULA DATA = error_dist;
VAR Res_RGDP Res_UMP Res_CPI;
FIT Normal /
MARGINALS = uniform METHOD = MLE;
SIMULATE /
NDRAWS = 1000 SEED = 1
MARGINALS = UNIFORM
OUTUNIFORM = simulated_errors;
RUN; The Error_Dist has values like these: Res_RGDP Res_UMP Res_CPI 0.0398668592 0.0262994594 0.010727359 -0.029316863 0.1554613968 0.0061404736 -0.050879246 0.0763711127 -0.00331784 0.0068502877 0.1468297946 0.0019326859 0.0650468136 -0.018832064 0.0044021124 -0.008798372 0.1423033571 0.0029771118
... View more
Rookie here. I am trying to evaluate the following. The idea is that this should evaluate lp with an initial value (given in the first equation) and then loop to populate with those fixed values (e.g. kA, kB) and array values of A, B, C values. Initial values are zero for array AX, BX and CX. When I run it, i get zeroes for all instances of A, B and C and no LP values. Any idea where i am wrong? data test1 (drop = i); array AX[128] A1-A128; array BX[128] B1-V128; array CX[128] C1-T128; do i = 1 to 128; AX[i] = 0; BX[i] = 0; CX[i] = 0; end; run; data test2 ;
set test1;
array AX[128] A1-A128;
array BX[128] B1-V128;
array CX[128] C1-T128;
kA = 23.99196;
kB = 21.04797;
kC = 54.98294;
lp = (log(kA) * 127/2) + (log(kB) * 127/2) + (log(kC) * 127/2);
do i=2 to 128;
lp = lp - kA * (AX[i] - AX[i-1])**2 / 2
- kB * (BX[i] -BX[i-1])**2 / 2
- kC * (CX[i] - CX[i-1])**2 / 2;
end;
run;
... View more
hi. The pattern is: 3 variable types: A, V, T. A ranges from 1-11, V ranges from 0-11 and T ranges from 1-23 (e.g. A1, A2, V0, V1, T1, T2, etc.). The values for these variables are 0,1.
... View more
thanks. that paper is helpful. however i am still getting an error (ideally stupidity from my end!). I am using the following argument: proc mcmc DATA=input4 diag=all dic
propcov=quanew
nbi=5000
ntu=5000
nmc=10000
thin=20
plots(smooth)=all seed=19680409
outpost=pdres;
array od[46] a1-a11 v0-v11 t1-t23;
array bd[46] ab_1-ab_11 vb_0-vb_11 tb_1-tb_23;
parms (ab_1-ab_11 vb_0-vb_11 tb_1-tb_23) 0;
parms sigma2 1;
/* Prior distribution assumptions */
prior sigma2 ~ igamma(shape = 100, scale= 76);
prior ab_1-ab_11 vb_0-vb_11 tb_1-tb_23 ~ normal(mean=0, var=1e6);
do i = 1 to 46;
p = od[i] * bd[i];
end;
model d ~ normal(p, var=sigma2);
run; The error says: All observations in the input data set contain missing values. Any idea where I am going wrong?
... View more
Rookie here and struggling to create a declaration to run regression for multiple independent variables. The dataset looks like the following A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11 V0 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 T13 T14 T15 T16 T17 T18 T19 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 As part of Proc MCMC, I want to specify the following a model in the following manner: y = a + a_b1*A1 + a_b2 *A2 + a_b3*A3... + b_b1*V1 + b_b2*V2 + ... and so on Is there a shortcut way of doing this for a large sets of A, V and T?
... View more
Thanks Chris. However, that code does not seem to be working. I have slightly edited the original post table. Ideally I am looking to populate the P_Current field with the last Current value (i.e. value from the previous row). With the code you suggested, the P_Current is populated with all 1s. Any idea where I might be going wrong?
... View more
I have a table where I need a new variable (P_Current) that takes value of the previous value of variable (Current) . The following highlights what I am looking for: ID Current P_Current 1 1 1 2 0 1 2 1 0 3 0 1 4 0 0 5 1 0 6 1 1 I was trying the following but thats probably not correct data test2;
set test1;
by ID;
P_Current = lag(Current);
if first.ID then P_Current = Current;
run;
... View more
Trying to understand how APC models work in terms of default of retail loan portfolio. I looked around but could not find much resource on working solutions. Does anyone have any pointers as to where I can find some material with SAS references? - Thanks
... View more
This is great! Thanks so much for the quick response. I had one small query. When I run the code, I notice that once an account has a cure, the subsequent cure values remain as 1 even when it has gone back to redefault. Any idea how I can reset the count so that once an account has a cure of 1, the next month, it will go back to zero as is the case with n_months.
... View more
SAS rookie here. i am trying to calculate default and redefault rate for a portfolio. I have the following logic to work through Cure is defined as an account being in good health for 6 months following a default.So if an account is in default(denoted as 1) at time 1, then if there are no more defaults till time 6, then at time 7, account is said to be 'Cured' (denoted as Cured = 1). Default is satisfied if one of the two flag variables are true (FirstDefaultFlag, DefaultFlag). An account can get multiple defaults and cures. I just need to flag them. I have the following code but obviously it is incorrect. data p1;
set p0;
by AccountID;
If DefaultFlag = 'Y' or FirstDefaultFlag = 'Y' then Default = 1;
else Default = 0;
if Default = 1 then do;
row = 0;
selected = 0;
cure = 0;
end;
row + 1;
retain selected;
if selected = 0 and row <= 6 and Default = 0 then do;
Cure = 1;
selected = 1;
end;
drop row selected;
run; Any helpers?
... View more
