search

bulik / ldsc

LD Score Regression (LDSC)
GNU General Public License v3.0
652 stars 344 forks source link

FloatingPointError: invalid value encountered in less_equal #174

Open dekim22114 opened 5 years ago

dekim22114 commented 5 years ago

Hello, I recieved this error message. How can I solve this problem?

Beginning analysis at Fri Nov 1 20:19:43 2019 Reading summary statistics from JIA_munge.sumstats.gz ... Read summary statistics for 104285 SNPs. Reading reference panel LD Score from eur_w_ld_chr/[1-22] ... Read reference panel LD Scores for 1290028 SNPs. Removing partitioned LD Scores with zero variance. Reading regression weight LD Score from eur_w_ld_chr/[1-22] ... Read regression weight LD Scores for 1290028 SNPs. After merging with reference panel LD, 48398 SNPs remain. After merging with regression SNP LD, 48398 SNPs remain. Computing rg for phenotype 2/2 Reading summary statistics from PASS/PASS_Type_1_Diabetes.anno.sumstats ... Read summary statistics for 915517 SNPs. After merging with summary statistics, 39528 SNPs remain. 39528 SNPs with valid alleles. WARNING: number of SNPs less than 200k; this is almost always bad. /home/de/fusion/fusion_twas-master/ldsc/ldscore/sumstats.py:532: FutureWarning: Method .as_matrix will be removed in a future version. Use .values instead. ref_ld = sumstats.as_matrix(columns=ref_ld_cnames)

ERROR computing rg for phenotype 2/2, from file PASS/PASS_Type_1_Diabetes.anno.sumstats. Traceback (most recent call last): File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/sumstats.py", line 410, in estimate_rg rghat = _rg(loop, args, log, M_annot, ref_ld_cnames, w_ld_cname, i) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/sumstats.py", line 539, in _rg intercept_gencov=intercepts[2], n_blocks=n_blocks, twostep=args.two_step) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/regressions.py", line 690, in init slow=slow, twostep=twostep) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/regressions.py", line 346, in init slow=slow, step1_ii=step1_ii, old_weights=old_weights) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/regressions.py", line 213, in init x, yp, update_func, n_blocks, slow=slow, w=initial_w) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/irwls.py", line 66, in init x, y, update_func, n_blocks, w, slow=slow, separators=separators) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/irwls.py", line 113, in irwls new_w = np.sqrt(update_func(cls.wls(x, y, w))) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/irwls.py", line 159, in wls x = cls._weight(x, w) File "/home/de/fusion/fusion_twas-master/ldsc/ldscore/irwls.py", line 187, in _weight if np.any(w <= 0): FloatingPointError: invalid value encountered in less_equal

AMSSwanglab commented 4 years ago

I got the same error... have you solved this problem?

bcrone commented 3 years ago

Also receiving this error - bumping for visibility

sandiassaf commented 3 years ago

I am encountering the same issue. Was anyone able to resolve this?

kllrdr commented 3 years ago

Hi all,

I got the same error. I could figure out that the problem is having NaN (Not a Number) in our data, which is a member of a numeric data type that can be interpreted as a value that is undefined or unrepresentable, especially in floating-point arithmetic. Has anyone solved this problem since posting?

chenming9453 commented 3 years ago

Hi all, I got the same error here. I am wondering if anyone has solved this problem?

kllrdr commented 3 years ago

Hi,

Yes, the problem is having missing values somewhere in your annotations (which you used to compute LD scores). So, some SNPs are missing some values for different partitions. Example:

      AN1 AN2 AN3 AN4

SNP1 0.34 0.23 0.45 0.12 SNP2 0.24 0.45 SNP3 0.12 0.46 0.23 0.34 SNP4 0.34 0.22 0.43

If you open your ldscore.gz files, you will see NaN values instead of LD scores for various annotations. I made my annotations again and the problem was solved.

On Mon, May 10, 2021 at 11:08 PM chenming9453 @.***> wrote:

Hi all, I got the same error here. I am wondering if anyone has solved this problem?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/bulik/ldsc/issues/174#issuecomment-837704488, or unsubscribe https://github.com/notifications/unsubscribe-auth/ATHGTPXLU6FS4KGPHSVYD6LTNCNRTANCNFSM4JHZOMJA .

goodluckncb commented 1 year ago

I also ran into this problem, I built the sumstats.gz file myself and needed to determine if there were nan, inf,-inf equivalents in the "Z" column, just make those values clear

isabellebond commented 1 year ago

Hi,

I've been using the files from eur_w_ld_chr/ and have run all my sumstats through munge_sumstats to get them in the correct format and am also getting this error. I've checked for missing data in any of the .ldscore files and haven't found anything. Does anyone know of anything else that might be causing this issue? Thank you so much

AndreaG5 commented 7 months ago

Hi all,

I am encountering the same problem. I checked for missing values in my *.sumstats.gz file, but it doesn't seem to be any of NA/Na/NaN. Does anyone solve the problem in any way?

Thanks