curvefit.core.data.Data
Manages all data to be used for curve fitting
The data class contains all information about the complete dataset, list of groups, column names.
Creates observation standard error based on obs_se_func
.
All sorting will happen immediately in the Data
class init
and we won’t have to sort again.
NOTE: will prioritize obs_se_func
over col_obs_se
.
Syntax
d = Data(df, col_t, col_obs, obs_space, col_covs, col_group)
Arguments
df (pandas.DataFrame)
: all datacol_t (str)
: column with timecol_obs (str)
: column with observation valuesobs_space (callable)
: function from functions module that specifies what space thecol_obs
column represents (e.g. ln_gaussian_cdf)col_covs (List[str])
: covariate columnscol_group (str)
: group columncol_obs_se (optional, str)
: column with observation standard error specified.obs_se_func (optional, callable)
: observation standard error function to create from time
Attributes
self.groups (List[str])
: list of groups, sorted in orderself.num_obs (int)
: dictionary of numbers of observations per group (key)self.times (Dict[str: np.array]))
: dictionary of times per group (key)
Methods
_get_df
Returns a copy of the data frame or a pointer to the data frame, and optionally data specs.
If you plan on modifying the data frame, use copy=True
. If a group is passed,
then a group-specific data frame will be passed.
group (optional, str)
: optional group namecopy (bool)
: return a copy or not
_get_translated_observations
Returns the observations for a group in a particular space from curvefit.core.functions
group (str)
: which group to return observations forspace (callable)
: which space to translate observations to