config¶
-
class
geocoding.config.Config[source]¶ Class that configures the execution process.
-
n_folds¶ The number of folds in the experiment
- Type
int
-
source_crs¶ The EPSG crs code used in input files
- Type
int
-
target_crs¶ The EPSG crs code to transform the data
- Type
int
-
clusters_pct¶ Percentage of data points, indicating how many clusters to create in order to query Overpass API for streets
- Type
float
-
osm_buffer¶ A buffer distance (in meters) to consider around each bounding box when querying Overpass API
- Type
float
-
osm_timeout¶ Timeout (in seconds) after five requests to Overpass API
- Type
int
-
max_overpass_tries¶ Maximum number of failed tries to extract the road network when querying the Overpass API before quiting.
- Type
int
-
distance_thr¶ Distances in features greater than this value will be converted to this threshold
- Type
float
-
baseline_service¶ The name of the service to consider when measuring baseline scores
- Type
str
-
experiments_path¶ Path to folder that stores the experiments
- Type
str
-
services¶ The services (geocoders) used in the setup
- Type
list
-
supported_features¶ List of the supported features to choose from
- Type
list
-
included_features¶ List of the features to be included in the experiment
- Type
list
-
normalized_features¶ List of features to be normalized
- Type
list
-
supported_classifiers¶ List of the supported classifiers to choose from
- Type
list
-
included_classifiers¶ List of the classifiers to be included in the experiment
- Type
list
-
NB_hparams¶ Parameters search space for Naive Bayes classifier
- Type
dict
-
NN_hparams¶ Parameters search space for Nearest Neighbors classifier
- Type
dict
-
LR_hparams¶ Parameters search space for Logistic Regression classifier
- Type
dict
-
SVM_hparams¶ Parameters search space for SVM classifier
- Type
list
-
MLP_hparams¶ Parameters search space for MLP classifier
- Type
dict
-
DT_hparams¶ Parameters search space for Decision Tree classifier
- Type
dict
-
RF_hparams¶ Parameters search space for Random Forest classifier
- Type
dict
-
ET_hparams¶ Parameters search space for Extra Trees classifier
- Type
dict
-
hyperparams_search_method= 'grid'¶ Search Method to use for finding best hyperparameters. (randomized | grid).
- Type
str
-
max_iter= 30¶ Number of iterations that RandomizedSearchCV should execute. It applies only when
hyperparams_search_methodequals to ‘randomized’.- Type
int
-
n_jobs= 4¶ Number of parallel jobs to be initiated. -1 means to utilize all available processors.
- Type
int
-
seed_no= 13¶ Seed to use by random number generators.
- Type
int
-
