https://github.com/gersteinlab/nb-fit

Negative binomial regression fitting code in C++. Adapted from R source code.
https://github.com/gersteinlab/nb-fit

Last synced: 2 months ago
JSON representation

Negative binomial regression fitting code in C++. Adapted from R source code.

• Host: GitHub
• URL: https://github.com/gersteinlab/nb-fit
• Owner: gersteinlab
• Created: 2018-01-31T17:27:38.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-01-31T17:29:42.000Z (over 7 years ago)
• Last Synced: 2025-01-20T12:07:47.119Z (4 months ago)
• Language: C++
• Size: 26.4 KB
• Stars: 0
• Watchers: 26
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.txt

Awesome Lists containing this project

README

        
Usage:

glm.nb [response file] [predictor file] [initial theta]

Inputs:

[response file]: A text file containing the responses (y vector) for each observation, one per line.

[predictor file]: A text file containing the feature values (x matrix) to use in the negative binomial fitting, where rows are observations, and columns are features.

[initial theta]: Used to determine convergence of the underlying algorithms. In R, this is typically set to 0.000001

Output:

"glm.nb" prints to standard output, which can be redirected. This output includes:

- Coefficients

- Residuals

- Fitted Values

- Effects

- R (correlation measure)

- Rank

- QR matrix

- Qraux

- Pivot vector

- Tol

- Linear predictors

- Deviance

- AIC

- Null deviance

- Number of iterations

- Weights

- Prior weights

- Degrees of freedom residual

- Degrees of freedom null

- Converged

- Boundary

- Theta

- SE Theta

- Two Log Likelihood