I can calculate the motion of heavenly bodies but not the madness of people. -Isaac Newton

## Ibis – A New Kid in Town

Developed by Wes McKinney, pandas is a very efficient and powerful data analysis tool in python language for data scientists. Same as R, pandas reads the data into memory. As a result, we might often face the problem of running out of memory while analyzing large-size data with pandas.

Similar to Blaze, ibis is a new data analysis framework in python built on top of other back-end data engines, such as sqlite and impala. Even better, ibis provides a higher compatibility to pandas and better performance than Blaze.

In a previous blog (https://statcompute.wordpress.com/2015/03/27/a-comparison-between-blaze-and-pandas), I’ve shown the efficiency of Blaze through a simple example. However, in the demonstration below, it is shown that, while applied to the same data with sqlite engine, ibis is 50% more efficient than Blaze in terms of the “real time”.

```import ibis as ibis

tbl = ibis.sqlite.connect('//home/liuwensui/Documents/data/flights.db').table('tbl2008')
exp = tbl[tbl.DayOfWeek > 1].group_by("DayOfWeek").aggregate(avg_AirTime = tbl.AirTime.mean())
pd = exp.execute()
print(pd)

#i   DayOfWeek  avg_AirTime
#0          2   103.214930
#1          3   103.058508
#2          4   103.467138
#3          5   103.557539
#4          6   107.400631
#5          7   104.864885
#
#real   0m10.346s
#user   0m9.585s
#sys    0m1.181s
```

Written by statcompute

November 15, 2015 at 12:56 am

## Quasi-Binomial Model in SAS

Similar to quasi-Poisson regressions, quasi-binomial regressions try to address the excessive variance by the inclusion of a dispersion parameter. In addition to addressing the over-dispersion, quasi-binomial regressions also demonstrate extra values in other areas, such as LGD model development in credit risk modeling, due to its flexible distributional assumption.

Measuring the ratio between NCO and GCO, LGD could take any value in the range [0, 1] with no unanimous consensus on the distributional assumption currently in the industry. An advantage of quasi-binomial regression is that it makes no assumption of a specific distribution but merely specifies the conditional mean for a given model response. As a result, the trade-off is the lack of likelihood-based measures such as AIC and BIC.

Below is a demonstration on how to estimate a quasi-binomial model with GLIMMIX procedure in SAS.

```proc glimmix data = _last_;
model y = age number start / link = logit solution;
_variance_ = _mu_ * (1-_mu_);
random _residual_;
run;
/*
Model Information
Data Set                     WORK.KYPHOSIS
Response Variable            y
Response Distribution        Unknown
Variance Function            _mu_ * (1-_mu_)
Variance Matrix              Diagonal
Estimation Technique         Quasi-Likelihood
Degrees of Freedom Method    Residual

Parameter Estimates
Standard
Effect       Estimate       Error       DF    t Value    Pr > |t|
Intercept     -2.0369      1.3853       77      -1.47      0.1455
age           0.01093    0.006160       77       1.77      0.0800
number         0.4106      0.2149       77       1.91      0.0598
start         -0.2065     0.06470       77      -3.19      0.0020
Residual       0.9132           .        .        .         .
*/
```

For the comparison purpose, the same model is also estimated with R glm() function, showing identical outputs.

```summary(glm(data = kyphosis, Kyphosis ~ ., family = quasibinomial))
#Coefficients:
#            Estimate Std. Error t value Pr(>|t|)
#(Intercept) -2.03693    1.38527  -1.470  0.14552
#Age          0.01093    0.00616   1.774  0.07996 .
#Number       0.41060    0.21489   1.911  0.05975 .
#Start       -0.20651    0.06470  -3.192  0.00205 **
#---
#(Dispersion parameter for quasibinomial family taken to be 0.913249)
```

Written by statcompute

November 1, 2015 at 1:20 pm