ValueError: The first guess on the deviance function returned a nan. This could be a boundary problem and should be reported. #20
Description
Hi,
I am trying to run STARRpeaker for the first time and I am encountering the following errors.
Any idea how I can solve it?
Thank you,
Ana
[2025-05-02 15:56:23] Loading fragment coverages
[2025-05-02 15:57:42] Loading covariates
[2025-05-02 15:58:37] Total genomic bins: 10,501,586
[2025-05-02 15:58:37] Total non-overlapping genomic bins: 2,100,332
[2025-05-02 15:58:59] Removing bins with input counts larger than 36.9735 for training and testing
[2025-05-02 15:58:59] Removing bins with output counts larger than 23,060.12842 for training
[2025-05-02 15:58:59] Total genomic bins used for training: 1
[2025-05-02 15:58:59] Total genomic bins used for testing: 0
[2025-05-02 15:58:59] Running Mode 1
[2025-05-02 15:58:59] Fit using formula: y~x1+x2
/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/core.py:505: RuntimeWarning: divide by zero encountered in true_divide
t0 = n / sum(weights * (y / mu - 1) ** 2)
/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/core.py:486: RuntimeWarning: invalid value encountered in subtract
return sum(w * (digamma(th + y) - digamma(th) + np.log(th) + 1 - np.log(th + mu) - (y + th) / (mu + th)))
/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/core.py:486: RuntimeWarning: invalid value encountered in true_divide
return sum(w * (digamma(th + y) - digamma(th) + np.log(th) + 1 - np.log(th + mu) - (y + th) / (mu + th)))
/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/core.py:490: RuntimeWarning: invalid value encountered in true_divide
return sum(w * (-trigamma(th + y) + trigamma(th) - 1 / th + 2 / (mu + th) - (y + th) / (mu + th) ** 2))
[2025-05-02 15:58:59] Initial estimate of theta is nan
Traceback (most recent call last):
File "/home/aa733/miniconda3/envs/starrpeaker/bin/starrpeaker", line 8, in
sys.exit(main())
File "/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/starrpeaker.py", line 84, in main
minFC=args.minfc)
File "/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/starrpeaker/core.py", line 686, in call_peak
model = smf.glm(formula, data=df, family=sm.families.NegativeBinomial(alpha=1 / th0)).fit(start_params=model0.params)
File "/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/statsmodels/genmod/generalized_linear_model.py", line 1028, in fit
cov_kwds=cov_kwds, use_t=use_t, **kwargs)
File "/home/aa733/miniconda3/envs/starrpeaker/lib/python2.7/site-packages/statsmodels/genmod/generalized_linear_model.py", line 1143, in _fit_irls
raise ValueError("The first guess on the deviance function "
ValueError: The first guess on the deviance function returned a nan. This could be a boundary problem and should be reported.