Error summary: HailException

Hail Query & hailctl
1

Hey,
I’m trying to run a logistic regression on my matrix table:

gwas = hl.logistic_regression_rows(test='wald',
                                  y=hl.int(mt_filtered.pheno["Sel Related Thrombocytopenia"]),
                                  x=mt_filtered.GT.n_alt_alleles(),
                                  covariates=cov)

p = hl.plot.manhattan(gwas.p_value)
show(p)

but, I keep getting this error:

Hail version: 0.2.82-2ab242915c2c
Error summary: HailException: expected 294 fields, but found 295 fields

I passed quote=’"’ to the import_table, but still with no change.

please help me :slight_smile:
thanks!
Shiri

2

This is a problem with table import, so the logistic regression isn’t an important part of debugging the pipeline. What’s the code you use to import your table?

Also, upgrading to the latest version would be good, because the error message might be slightly different.

3

Hi,
the import table code is:

table = hl.import_table("meta_data/023-BOSTON_samples_edited.csv", impute=True, key="USUBJID", delimiter=',', quote='"')

I filled all the null etries with 0. Now I’m getting this error:
Error summary: HailException: No complete samples: each sample is missing its phenotype or some covariate

What could be the problem?..

4

Is this issue resolved? I’m facing this issue too.

5

This error occurs when no samples have a defined value for the phenotype and each covariate. Take a look at missingness patterns of those fields / inputs – there’s probably a lot more missingness than you expect.