> I have recently tried using the quantiles_summary and qp_plot modules but
> have run into some roadblocks.  I first ran a model called 'model' then
> called

> quantiles_summary(model, method='deviance', n_samples=50,
>                       quantiles = (10, 30, 50, 70, 90), sorted_idx = None,
>                       cdf_range  = (-6, 6))

> Which led kicked back:

> AttributeError                            Traceback (most recent call last)
> <ipython-input-43-8a6de8a1d14d> in <module>()
>       1 quantiles_summary(model, method='deviance', n_samples=50,
>       2                       quantiles = (10, 30, 50, 70, 90), sorted_idx =
> None,
> ----> 3                       cdf_range  = (-5, 5))

> /Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packag es/hddm/utils.pyc
> in quantiles_summary(hm, method, n_samples, quantiles, sorted_idx,
> cdf_range)
>     636     #Init
>     637     n_q = len(quantiles)
> --> 638     wfpt_dict = hm.params_dict['wfpt'].subj_nodes
>     639     conds = wfpt_dict.keys()
>     640     n_conds = len(conds)

> AttributeError: 'HDDM' object has no attribute 'params_dict'

> Do I need to input the models parameter dict myself somewhere? After getting
> a quantiles summary, I would like to create a quantile-probability plot of
> the data.  With respect to the qp_plot() function, I'm not exactly sure what
> the split_function does. Could someone provide an example of how this should
> be set up?  I have experimented with simulating data sets using the group
> estimated parameters but I would like to try this method out as well.

> Thank you for your help,

> Kyle

> Hi Kyle,

> Thanks for reporting, that's a bug. I filed an issue here:

> https://github.com/hddm-devs/hddm/issues/9

> Thomas

> On Tue, Sep 25, 2012 at 10:10 PM, DunovanK <duno...@gmail.com<javascript:>>
> wrote:
> > Hello all,

> > I have recently tried using the quantiles_summary and qp_plot modules
> but
> > have run into some roadblocks.  I first ran a model called 'model' then
> > called

> > quantiles_summary(model, method='deviance', n_samples=50,
> >                       quantiles = (10, 30, 50, 70, 90), sorted_idx =
> None,
> >                       cdf_range  = (-6, 6))

> > Which led kicked back:

> > AttributeError                            Traceback (most recent call
> last)
> > <ipython-input-43-8a6de8a1d14d> in <module>()
> >       1 quantiles_summary(model, method='deviance', n_samples=50,
> >       2                       quantiles = (10, 30, 50, 70, 90),
> sorted_idx =
> > None,
> > ----> 3                       cdf_range  = (-5, 5))

> /Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packag es/hddm/utils.pyc

> > in quantiles_summary(hm, method, n_samples, quantiles, sorted_idx,
> > cdf_range)
> >     636     #Init
> >     637     n_q = len(quantiles)
> > --> 638     wfpt_dict = hm.params_dict['wfpt'].subj_nodes
> >     639     conds = wfpt_dict.keys()
> >     640     n_conds = len(conds)

> > AttributeError: 'HDDM' object has no attribute 'params_dict'

> > Do I need to input the models parameter dict myself somewhere? After
> getting
> > a quantiles summary, I would like to create a quantile-probability plot
> of
> > the data.  With respect to the qp_plot() function, I'm not exactly sure
> what
> > the split_function does. Could someone provide an example of how this
> should
> > be set up?  I have experimented with simulating data sets using the
> group
> > estimated parameters but I would like to try this method out as well.

> > Thank you for your help,

> > Kyle