First attempt at Quantile probability plot #319
Conversation
| ).reshape(out_shape) | ||
| lapse_response = rng.binomial(n=1, p=0.5, size=out_shape) | ||
| ).reshape(replace_shape) | ||
| lapse_response = rng.binomial(n=1, p=0.5, size=replace_shape) |
There was a problem hiding this comment.
here I would go with np.random.choice(n_choices, ...) to accomodate models that have more than 2 choices.
if n_choices == 2, we change 0 --> -1, otherwise keep choices as they are.
There was a problem hiding this comment.
Can we tab this for now since it might take a bit more engineering to figure out where that n_choices comes from. ssms would need that to figure out how many responses to generate right?
There was a problem hiding this comment.
yeah. That should be part of model configs actually so should be easily accessible. (can double check but I think it is for ssm-simulators?)
There was a problem hiding this comment.
Also outputs from ssms.simulator() have the 'metadata' key, which also has that.
There was a problem hiding this comment.
I mean we do have one step to figure out number of responses from the data, but we might want the user to be a bit more explicit in non-binary cases. In any case we might need a separate PR to address this
| _logger = logging.getLogger("hssm") | ||
|
|
||
|
|
||
| def _plot_quantile_probability_1D( |
There was a problem hiding this comment.
some of these plotting parameters might be passed through to the plot as **kwargs?
(I know I didn't do this well in HDDM :))
There was a problem hiding this comment.
reading a bit more I think you already tried to account for that quite a bit, but took the choice e.g. to keep xlabel and ylabel as named arguments here?
There was a problem hiding this comment.
Sorry just saw this. Yeah I am doing xlabel and ylabel from experience with R lol. It's hard to decide which ones you explicitly declare and which ones can be included in kwargs. I just thought that the users might want to be able to change these labels quickly
|
|
||
| # get the posterior samples | ||
| idata_posterior = idata["posterior_predictive"]["rt,response"] |
There was a problem hiding this comment.
seems like logic wise, you could have idata is None --> True but then still access idata in this line?
There was a problem hiding this comment.
The function returns if idata is None at line 121, so it would not reach here if idata is None?
There was a problem hiding this comment.
ok overlooked the return
|
|
||
| # get the posterior samples | ||
| idata_posterior = idata["posterior_predictive"]["rt,response"] |
There was a problem hiding this comment.
ok overlooked the return
I have done the following in this PR:
seabornto basicmatplotlibplots with the ability to plot regions of uncertainty