model.vma
Implementation of the Variable Meta-Analysis module.
def
convert_percentages(val):
pd.apply() functions to convert percentages
def
convert_NaN(val):
pd.apply() functions to convert NaN
def
normalize_units(val):
class
VMA:
Meta-analysis of multiple data sources to a summary result.
VMA( filename, title=None, low_sd=1.0, high_sd=1.0, discard_multiplier=3, stat_correction=None, use_weight=False, bound_correction=None, description=None, units=None)
Arguments: filename: (string, pathlib.Path, or io.StringIO) Can be either
- Path to a CSV file containing data sources. The CSV file must contain columns named "Raw Data Input", "Weight", and "Original Units". It can contain additional columns, which will be ignored.
- io.StringIO objects are processed as if they are opened CSV files title: string, name of the VMA to extract from an Excel file. This value is unused if filename is a CSV. Will raise an AssertionError if the title is not available in the Excel file. low_sd: number of multiples of the stddev to use for the low result. high_sd: number of multiples of the stddev to use for the high result. discard_multiplier: discard outlier values more than this many multiples of the stddev away from the mean. stat_correction: discard outliers more than discard_multiplier stddev away from the mean. use_weight: if true, use weights provided with the VMA to bias the mean. bound_correction: if true, and the low value calculated with standard deviation would be negative, use min instead of sd on on the lower value. description: optional description of what this VMA describes units: the units to use for this VMA; retrieved from datafile by default
def
avg_high_low( self, key=None, regime=None, region=None, low_sd=None, high_sd=None, discard_multiplier=None, stat_correction=None, use_weight=None, bound_correction=None):
Args
- key: (optional) specify 'mean', 'high' or 'low' to get single value
- regime: string name of the thermal moisture regime to select sources for.
- region: string name of the world region to select sources for.
- Other parameters: explicitly override the default parameters for this VMA.
Returns
By default returns (mean, high, low) using low_sd/high_sd. If key is specified will return associated value only
def
essential_parameters(self):
Return a dictionary of "essential" parameters, that is parameters whose value differs from the default