Hi All,
I conducted a regression analysis on data from over 3,000 U.S. counties, covering the period from 2016 to 2020. However, when I executed the xsmle command, I repeatedly encountered a *3200 conformability error. My five-year panel dataset comprises 15,445 observations, and the associated matrix has dimensions of 15,445 by 15,445. Furthermore, my dataset is strongly balanced. Below is the Stata code I used:
use "$raw_data/spatial_matrix/usc_spatial_weight.dta", clear
rename fips FIPS
// Drop Alaska and Hawaii
drop if state == "02" | state == "15"
destring FIPS, replace
// The following counties are not included in the dataset used for regression, so they are excluded from the spatial weight matrix. They are removed from both rows and columns:
drop if inlist(FIPS, 46113, 51515, 51520, 51560, 51580, 51590, 51595, 51600, 51630, 51678, 51685, 51720, 51730, 51820, 51830, 51840, 25007, 25019, 36085, 53055)
drop v46113 v51515 v51520 v51560 v51580 v51590 v51595 v51600 v51630 v51678 v51685 v51720 v51730 v51820 v51830 v51840 v25007 v25019 v36085 v53055
keep v*
save "$temp_data/weight1.dta", replace
spatwmat using "$temp_data/weight1.dta", n(W)
// Repeat the matrix for five years to align with the five-year panel data
spcs2xt v*, matrix(W5) time(5)
// This is the data used for regression
use "$working_data/mw_county_reg_full_sample.dta", clear
drop if Statefip == 2 | Statefip == 15
keep if year >= 2016
xtset FIPS year
sort year FIPS
spatwmat using "$temp_data/W5xt.dta", name(Wxt) standardize
xsmle y x control, re model(sar) wmat(Wxt) type(both) nolog noeffects
Results:
Warning: Option type(both) will be ignored
*: 3200 conformability error
_xsmle_est(): - function returned error
<istmt>: - function returned error
Any suggestions? Any help would be much appreciated.
I conducted a regression analysis on data from over 3,000 U.S. counties, covering the period from 2016 to 2020. However, when I executed the xsmle command, I repeatedly encountered a *3200 conformability error. My five-year panel dataset comprises 15,445 observations, and the associated matrix has dimensions of 15,445 by 15,445. Furthermore, my dataset is strongly balanced. Below is the Stata code I used:
use "$raw_data/spatial_matrix/usc_spatial_weight.dta", clear
rename fips FIPS
// Drop Alaska and Hawaii
drop if state == "02" | state == "15"
destring FIPS, replace
// The following counties are not included in the dataset used for regression, so they are excluded from the spatial weight matrix. They are removed from both rows and columns:
drop if inlist(FIPS, 46113, 51515, 51520, 51560, 51580, 51590, 51595, 51600, 51630, 51678, 51685, 51720, 51730, 51820, 51830, 51840, 25007, 25019, 36085, 53055)
drop v46113 v51515 v51520 v51560 v51580 v51590 v51595 v51600 v51630 v51678 v51685 v51720 v51730 v51820 v51830 v51840 v25007 v25019 v36085 v53055
keep v*
save "$temp_data/weight1.dta", replace
spatwmat using "$temp_data/weight1.dta", n(W)
// Repeat the matrix for five years to align with the five-year panel data
spcs2xt v*, matrix(W5) time(5)
// This is the data used for regression
use "$working_data/mw_county_reg_full_sample.dta", clear
drop if Statefip == 2 | Statefip == 15
keep if year >= 2016
xtset FIPS year
sort year FIPS
spatwmat using "$temp_data/W5xt.dta", name(Wxt) standardize
xsmle y x control, re model(sar) wmat(Wxt) type(both) nolog noeffects
Results:
- Weight Matrix:
- Imported binary weights matrix Wxt
- Dimension: 15,445 x 15,445
- Imported binary weights matrix Wxt
- Strongly Balanced Data:
- xtset FIPS year
- Panel variable: FIPS (strongly balanced)
- Time variable: year, 2016 to 2020
- Delta: 1 unit
- Regression Error
Warning: Option type(both) will be ignored
*: 3200 conformability error
_xsmle_est(): - function returned error
<istmt>: - function returned error
Any suggestions? Any help would be much appreciated.
Comment