I'm trying to run hdfeppml on a panel of international trade flows. "Value" is my dv, "UKFTA" is my iv, "Expcom" and "Expmon" are manually created factor variables which I would like to use for fixed effects.
I follow online examples / help and get the error message:
"Error in stats::lm.wfit(x_resid, z_resid, mu) : 0 (non-NA) cases
In addition: Warning messages:
1: In mean.default(y) : argument is not numeric or logical: returning NA
2: In mean.default(y) : argument is not numeric or logical: returning NA"
I also tried to create a reprex for convenience and got another error message. All of this is outlined in the screenshot below. I've been on multiple forums and looked at several different approaches to both problems. Any help would be massively appreciated - thanks
Do not post an image of the output. Copy it from the console and paste it into your response, placing lines with three back ticks just before and after the output, like this:
```
pasted output of summary() goes here
```
Importer Exporter Commodity Month
Length:142780 Length:142780 Length:142780 Length:142780
Class :character Class :character Class :character Class :character
Mode :character Mode :character Mode :character Mode :character
Value UKFTA Expcom Expmon
Min. : 0.0 Min. :0.00000 Min. :10001 Min. :1010
1st Qu.: 15.7 1st Qu.:0.00000 1st Qu.:12612 1st Qu.:1235
Median : 387.2 Median :0.00000 Median :14846 Median :1462
Mean : 8166.4 Mean :0.01055 Mean :14985 Mean :1482
3rd Qu.: 2668.2 3rd Qu.:0.00000 3rd Qu.:17582 3rd Qu.:1734
Max. :2012001.3 Max. :1.00000 Max. :19988 Max. :1968
None of your variables are non-numeric, so I can't explain why you are getting a warning about that. Even Expcom and Expmon are numeric, though you describe them as factors.
Your call to hdfeppml() ends with fixed = "Expcom, "Expmon") that causes the fixed argument to be set to "Expcom" and cluster, which is the next argument, to be set to "Expmon". Is that what you want? If you want both columns to be passed to fixed use `fixed = c("Expcom, "Expmon"))
Thanks for this. I've changed my code accordingly but the error persists:
> Reg <- hdfeppml(data = Panel_long, dep = "Value", indep = "UKFTA", fixed = c("Expcom", "Expmon") )
Error in stats::lm.wfit(x_resid, z_resid, mu) : 0 (non-NA) cases
In addition: Warning messages:
1: In mean.default(y) : argument is not numeric or logical: returning NA
2: In mean.default(y) : argument is not numeric or logical: returning NA
I've also double checked all depends and imports. Any further advice on this would be great? Happy to share data if it helps.