megadetector 10.0.15__py3-none-any.whl

megadetector/utils/wi_taxonomy_utils.py

@@ -0,0 +1,1766 @@
+"""
+
+wi_taxonomy_utils.py
+
+Functions related to working with the SpeciesNet / Wildlife Insights taxonomy.
+
+"""
+
+#%% Imports and constants
+
+import os
+import json
+
+import pandas as pd
+
+from copy import deepcopy
+from collections import defaultdict
+from tqdm import tqdm
+
+from megadetector.utils.path_utils import \
+    insert_before_extension, find_images
+
+from megadetector.utils.ct_utils import (
+    split_list_into_n_chunks,
+    round_floats_in_nested_dict,
+    is_list_sorted,
+    invert_dictionary,
+    sort_list_of_dicts_by_key,
+    sort_dictionary_by_value,
+)
+
+from megadetector.postprocessing.validate_batch_results import \
+    validate_batch_results, ValidateBatchResultsOptions
+
+from megadetector.detection.run_detector import DEFAULT_DETECTOR_LABEL_MAP
+
+md_category_id_to_name = DEFAULT_DETECTOR_LABEL_MAP
+md_category_name_to_id = invert_dictionary(md_category_id_to_name)
+
+blank_prediction_string = \
+    'f1856211-cfb7-4a5b-9158-c0f72fd09ee6;;;;;;blank'
+no_cv_result_prediction_string = \
+    'f2efdae9-efb8-48fb-8a91-eccf79ab4ffb;no cv result;no cv result;no cv result;no cv result;no cv result;no cv result'
+animal_prediction_string = \
+    '1f689929-883d-4dae-958c-3d57ab5b6c16;;;;;;animal'
+human_prediction_string = \
+    '990ae9dd-7a59-4344-afcb-1b7b21368000;mammalia;primates;hominidae;homo;sapiens;human'
+vehicle_prediction_string = \
+    'e2895ed5-780b-48f6-8a11-9e27cb594511;;;;;;vehicle'
+
+non_taxonomic_prediction_strings = [blank_prediction_string,
+                                    no_cv_result_prediction_string,
+                                    animal_prediction_string,
+                                    vehicle_prediction_string]
+
+non_taxonomic_prediction_short_strings = [';'.join(s.split(';')[1:-1]) for s in \
+                                          non_taxonomic_prediction_strings]
+
+# Ignore some files when generating instances.json from a folder
+default_tokens_to_ignore = ['$RECYCLE.BIN']
+
+
+#%% Miscellaneous taxonomy support functions
+
+def is_valid_prediction_string(s):
+    """
+    Determine whether [s] is a valid WI prediction string.  Prediction strings look like:
+
+    '90d950db-2106-4bd9-a4c1-777604c3eada;mammalia;rodentia;;;;rodent'
+
+    Args:
+        s (str): the string to be tested for validity
+
+    Returns:
+        bool: True if this looks more or less like a WI prediction string
+    """
+
+    # Note to self... don't get tempted to remove spaces here; spaces are used
+    # to indicate subspecies.
+    return isinstance(s,str) and (len(s.split(';')) == 7) and (s == s.lower())
+
+
+def is_valid_taxonomy_string(s):
+    """
+    Determine whether [s] is a valid 5-token WI taxonomy string.  Taxonomy strings
+    look like:
+
+    'mammalia;rodentia;;;;rodent'
+    'mammalia;chordata;canidae;canis;lupus dingo'
+
+    Args:
+        s (str): the string to be tested for validity
+
+    Returns:
+        bool: True if this looks more or less like a WI taxonomy string
+    """
+    return isinstance(s,str) and (len(s.split(';')) == 5) and (s == s.lower())
+
+
+def clean_taxonomy_string(s, truncate_multiple_description_strings=True):
+    """
+    If [s] is a seven-token prediction string, trim the GUID and common name to produce
+    a "clean" taxonomy string.  Else if [s] is a five-token string, return it.  Else error.
+
+    Args:
+        s (str): the seven- or five-token taxonomy/prediction string to clean
+        truncate_multiple_description_strings (bool, optional): we use | to delimit
+            multiple descriptions in the same string; if this is True, clean and
+            return just the first, else error.
+
+    Returns:
+        str: the five-token taxonomy string
+    """
+
+    if truncate_multiple_description_strings:
+        tokens = s.split('|')
+        s = tokens[0]
+
+    if is_valid_taxonomy_string(s):
+        return s
+    elif is_valid_prediction_string(s):
+        tokens = s.split(';')
+        assert len(tokens) == 7
+        return ';'.join(tokens[1:-1])
+    else:
+        raise ValueError('Invalid taxonomy string')
+
+
+taxonomy_level_names = \
+    ['non-taxonomic','kingdom','phylum','class','order','family','genus','species','subspecies']
+
+
+def taxonomy_level_to_string(k):
+    """
+    Maps taxonomy level indices (0 for kindgom, 1 for phylum, etc.) to strings.
+
+    Args:
+        k (int): taxonomy level index
+
+    Returns:
+        str: taxonomy level string
+    """
+
+    assert k >= 0 and k < len(taxonomy_level_names), \
+        'Illegal taxonomy level index {}'.format(k)
+
+    return taxonomy_level_names[k]
+
+
+def taxonomy_level_string_to_index(s):
+    """
+    Maps strings ('kingdom', 'species', etc.) to level indices.
+
+    Args:
+        s (str): taxonomy level string
+
+    Returns:
+        int: taxonomy level index
+    """
+
+    assert s in taxonomy_level_names, 'Unrecognized taxonomy level string {}'.format(s)
+    return taxonomy_level_names.index(s)
+
+
+def taxonomy_level_index(s):
+    """
+    Returns the taxonomy level up to which [s] is defined (0 for non-taxnomic, 1 for kingdom,
+    2 for phylum, etc.  Empty strings and non-taxonomic strings are treated as level 0.  1 and 2
+    will never be returned; "animal" doesn't look like other taxonomic strings, so here we treat
+    it as non-taxonomic.
+
+    Args:
+        s (str): 5-token or 7-token taxonomy string
+
+    Returns:
+        int: taxonomy level
+    """
+
+    if s in non_taxonomic_prediction_strings or s in non_taxonomic_prediction_short_strings:
+        return 0
+
+    tokens = s.split(';')
+    assert len(tokens) in (5,7)
+
+    if len(tokens) == 7:
+        tokens = tokens[1:-1]
+
+    # Anything without a class is considered non-taxonomic
+    if len(tokens[0]) == 0:
+        return 0
+
+    # WI taxonomy strings start at class, so we'll never return 1 (kingdom) or 2 (phylum)
+    elif len(tokens[1]) == 0:
+        return 3
+    elif len(tokens[2]) == 0:
+        return 4
+    elif len(tokens[3]) == 0:
+        return 5
+    elif len(tokens[4]) == 0:
+        return 6
+    # Subspecies are delimited with a space
+    elif ' ' not in tokens[4]:
+        return 7
+    else:
+        return 8
+
+
+def is_taxonomic_prediction_string(s):
+    """
+    Determines whether [s] is a classification string that has taxonomic properties; this
+    does not include, e.g., blanks/vehicles/no cv result.  It also excludes "animal".
+
+    Args:
+        s (str): a five- or seven-token taxonomic string
+
+    Returns:
+        bool: whether [s] is a taxonomic category
+    """
+
+    return (taxonomy_level_index(s) > 0)
+
+
+
+def get_kingdom(prediction_string):
+    """
+    Return the kingdom field from a WI prediction string
+
+    Args:
+        prediction_string (str): a string in the semicolon-delimited prediction string format
+
+    Returns:
+        str: the kingdom field from the input string
+    """
+    tokens = prediction_string.split(';')
+    assert is_valid_prediction_string(prediction_string)
+    return tokens[1]
+
+
+def is_human_classification(prediction_string):
+    """
+    Determines whether the input string represents a human classification, which includes a variety
+    of common names (hiker, person, etc.)
+
+    Args:
+        prediction_string (str): a string in the semicolon-delimited prediction string format
+
+    Returns:
+        bool: whether this string corresponds to a human category
+    """
+    return prediction_string == human_prediction_string or 'homo;sapiens' in prediction_string
+
+
+def is_vehicle_classification(prediction_string):
+    """
+    Determines whether the input string represents a vehicle classification.
+
+    Args:
+        prediction_string (str): a string in the semicolon-delimited prediction string format
+
+    Returns:
+        bool: whether this string corresponds to the vehicle category
+    """
+    return prediction_string == vehicle_prediction_string
+
+
+def is_animal_classification(prediction_string):
+    """
+    Determines whether the input string represents an animal classification, which excludes, e.g.,
+    humans, blanks, vehicles, unknowns
+
+    Args:
+        prediction_string (str): a string in the semicolon-delimited prediction string format
+
+    Returns:
+        bool: whether this string corresponds to an animal category
+    """
+
+    if prediction_string == animal_prediction_string:
+        return True
+    if prediction_string == human_prediction_string or 'homo;sapiens' in prediction_string:
+        return False
+    if prediction_string == blank_prediction_string:
+        return False
+    if prediction_string == no_cv_result_prediction_string:
+        return False
+    if len(get_kingdom(prediction_string)) == 0:
+        return False
+    return True
+
+
+def taxonomy_info_to_taxonomy_string(taxonomy_info, include_taxon_id_and_common_name=False):
+    """
+    Convert a taxonomy record in dict format to a five- or seven-token semicolon-delimited string
+
+    Args:
+        taxonomy_info (dict): dict in the format stored in, e.g., taxonomy_string_to_taxonomy_info
+        include_taxon_id_and_common_name (bool, optional): by default, this function returns a
+            five-token string of latin names; if this argument is True, it includes the leading
+            (GUID) and trailing (common name) tokens
+
+    Returns:
+        str: string in the format used as keys in, e.g., taxonomy_string_to_taxonomy_info
+    """
+    s = taxonomy_info['class'] + ';' + \
+        taxonomy_info['order'] + ';'  + \
+        taxonomy_info['family'] + ';'  + \
+        taxonomy_info['genus'] + ';'  + \
+        taxonomy_info['species']
+
+    if include_taxon_id_and_common_name:
+        s = taxonomy_info['taxon_id'] + ';' + s + ';' + taxonomy_info['common_name']
+
+    return s
+
+
+#%% Functions used to manipulate results files
+
+def generate_whole_image_detections_for_classifications(classifications_json_file,
+                                                        detections_json_file,
+                                                        ensemble_json_file=None,
+                                                        ignore_blank_classifications=True,
+                                                        verbose=True):
+    """
+    Given a set of classification results in SpeciesNet format that were likely run on
+    already-cropped images, generate a file of [fake] detections in SpeciesNet format in which each
+    image is covered in a single whole-image detection.
+
+    Args:
+        classifications_json_file (str): SpeciesNet-formatted file containing classifications
+        detections_json_file (str): SpeciesNet-formatted file to write with detections
+        ensemble_json_file (str, optional): SpeciesNet-formatted file to write with detections
+            and classfications
+        ignore_blank_classifications (bool, optional): use non-top classifications when
+            the top classification is "blank" or "no CV result"
+        verbose (bool, optional): enable additional debug output
+
+    Returns:
+        dict: the contents of [detections_json_file]
+    """
+
+    with open(classifications_json_file,'r') as f:
+        classification_results = json.load(f)
+    predictions = classification_results['predictions']
+
+    output_predictions = []
+    ensemble_predictions = []
+
+    # i_prediction = 0; prediction = predictions[i_prediction]
+    for i_prediction,prediction in enumerate(predictions):
+
+        output_prediction = {}
+        output_prediction['filepath'] = prediction['filepath']
+        i_score = 0
+
+        if ignore_blank_classifications:
+
+            while (prediction['classifications']['classes'][i_score] in \
+                   (blank_prediction_string,no_cv_result_prediction_string)):
+
+                i_score += 1
+                if (i_score >= len(prediction['classifications']['classes'])):
+
+                    if verbose:
+
+                        print('Ignoring blank classifications, but ' + \
+                              'image {} has no non-blank values'.format(
+                                i_prediction))
+
+                    # Just use the first one
+                    i_score = 0
+                    break
+
+                # ...if we passed the last prediction
+
+            # ...iterate over classes within this prediction
+
+        # ...if we're supposed to ignore blank classifications
+
+        top_classification = prediction['classifications']['classes'][i_score]
+        top_classification_score = prediction['classifications']['scores'][i_score]
+        if is_animal_classification(top_classification):
+            category_name = 'animal'
+        elif is_human_classification(top_classification):
+            category_name = 'human'
+        else:
+            category_name = 'vehicle'
+
+        if category_name == 'human':
+            md_category_name = 'person'
+        else:
+            md_category_name = category_name
+
+        output_detection = {}
+        output_detection['label'] = category_name
+        output_detection['category'] = md_category_name_to_id[md_category_name]
+        output_detection['conf'] = 1.0
+        output_detection['bbox'] = [0.0, 0.0, 1.0, 1.0]
+        output_prediction['detections'] = [output_detection]
+        output_predictions.append(output_prediction)
+
+        ensemble_prediction = {}
+        ensemble_prediction['filepath'] = prediction['filepath']
+        ensemble_prediction['detections'] = [output_detection]
+        ensemble_prediction['prediction'] = top_classification
+        ensemble_prediction['prediction_score'] = top_classification_score
+        ensemble_prediction['prediction_source'] = 'fake_ensemble_file_utility'
+        ensemble_prediction['classifications'] = prediction['classifications']
+        ensemble_predictions.append(ensemble_prediction)
+
+    # ...for each image
+
+    ## Write output
+
+    if ensemble_json_file is not None:
+
+        ensemble_output_data = {'predictions':ensemble_predictions}
+        with open(ensemble_json_file,'w') as f:
+            json.dump(ensemble_output_data,f,indent=1)
+        _ = validate_predictions_file(ensemble_json_file)
+
+    output_data = {'predictions':output_predictions}
+    with open(detections_json_file,'w') as f:
+        json.dump(output_data,f,indent=1)
+    return validate_predictions_file(detections_json_file)
+
+# ...def generate_whole_image_detections_for_classifications(...)
+
+
+def generate_md_results_from_predictions_json(predictions_json_file,
+                                              md_results_file=None,
+                                              base_folder=None,
+                                              max_decimals=5,
+                                              convert_human_to_person=True,
+                                              convert_homo_species_to_human=True,
+                                              verbose=False):
+    """
+    Generate an MD-formatted .json file from a predictions.json file, generated by the
+    SpeciesNet ensemble.  Typically, MD results files use relative paths, and predictions.json
+    files use absolute paths, so this function optionally removes the leading string
+    [base_folder] from all file names.
+
+    Uses the classification from the "prediction" field if it's available, otherwise
+    uses the "classifications" field.
+
+    When using the "prediction" field, records the top class in the "classifications" field to
+    a field in each image called "top_classification_common_name".  This is often different
+    from the value of  the "prediction" field.
+
+    speciesnet_to_md.py is a command-line driver for this function.
+
+    Args:
+        predictions_json_file (str): path to a predictions.json file, or a dict
+        md_results_file (str, optional): path to which we should write an MD-formatted .json file
+        base_folder (str, optional): leading string to remove from each path in the
+            predictions.json file
+        max_decimals (int, optional): number of decimal places to which we should round
+            all values
+        convert_human_to_person (bool, optional): WI predictions.json files sometimes use the
+            detection category "human"; MD files usually use "person".  If True, switches "human"
+            to "person".
+        convert_homo_species_to_human (bool, optional): the ensemble often rolls human predictions
+            up to "homo species", which isn't wrong, but looks odd.  This forces these back to
+            "homo sapiens".
+        verbose (bool, optional): enable additional debug output
+
+    Returns:
+        dict: results in MD format
+    """
+
+    # Read predictions file
+    if isinstance(predictions_json_file,str):
+        with open(predictions_json_file,'r') as f:
+            predictions = json.load(f)
+    else:
+        assert isinstance(predictions_json_file,dict)
+        predictions = predictions_json_file
+
+    # Round floating-point values (confidence scores, coordinates) to a
+    # reasonable number of decimal places
+    if (max_decimals is not None) and (max_decimals > 0):
+        round_floats_in_nested_dict(predictions, decimal_places=max_decimals)
+
+    predictions = predictions['predictions']
+    assert isinstance(predictions,list)
+
+    # Convert backslashes to forward slashes in both filenames and the base folder string
+    for im in predictions:
+        im['filepath'] = im['filepath'].replace('\\','/')
+    if base_folder is not None:
+        base_folder = base_folder.replace('\\','/')
+
+    detection_category_id_to_name = {}
+    classification_category_name_to_id = {}
+
+    # Keep track of detections that don't have an assigned detection category; these
+    # are fake detections we create for non-blank images with non-empty detection lists.
+    # We need to go back later and give them a legitimate detection category ID.
+    all_unknown_detections = []
+
+    # Create the output images list
+    images_out = []
+
+    base_folder_replacements = 0
+
+    # im_in = predictions[0]
+    for im_in in predictions:
+
+        im_out = {}
+
+        fn = im_in['filepath']
+        if base_folder is not None:
+            if fn.startswith(base_folder):
+                base_folder_replacements += 1
+                fn = fn.replace(base_folder,'',1)
+
+        im_out['file'] = fn
+
+        if 'failures' in im_in:
+
+            im_out['failure'] = str(im_in['failures'])
+            im_out['detections'] = None
+
+        else:
+
+            im_out['detections'] = []
+
+            if 'detections' in im_in:
+
+                if len(im_in['detections']) == 0:
+                    im_out['detections'] = []
+                else:
+                    # det_in = im_in['detections'][0]
+                    for det_in in im_in['detections']:
+                        det_out = {}
+                        if det_in['category'] in detection_category_id_to_name:
+                            assert detection_category_id_to_name[det_in['category']] == det_in['label']
+                        else:
+                            detection_category_id_to_name[det_in['category']] = det_in['label']
+                        det_out = {}
+                        for s in ['category','conf','bbox']:
+                            det_out[s] = det_in[s]
+                        im_out['detections'].append(det_out)
+
+            # ...if detections are present
+
+            class_to_assign = None
+            class_confidence = None
+            top_classification_common_name = None
+
+            if 'classifications' in im_in:
+
+                classifications = im_in['classifications']
+                assert len(classifications['scores']) == len(classifications['classes'])
+                assert is_list_sorted(classifications['scores'],reverse=True)
+                class_to_assign = classifications['classes'][0]
+                class_confidence = classifications['scores'][0]
+
+                tokens = class_to_assign.split(';')
+                assert len(tokens) == 7
+                top_classification_common_name = tokens[-1]
+                if len(top_classification_common_name) == 0:
+                    top_classification_common_name = 'undefined'
+
+            if 'prediction' in im_in:
+
+                class_to_assign = None
+                im_out['top_classification_common_name'] = top_classification_common_name
+                class_to_assign = im_in['prediction']
+                if convert_homo_species_to_human and class_to_assign.endswith('homo species'):
+                    class_to_assign = human_prediction_string
+                class_confidence = im_in['prediction_score']
+
+            if class_to_assign is not None:
+
+                if class_to_assign == blank_prediction_string:
+
+                    # We have a blank prediction with detections present.  For now, don't do anything
+                    # special here, just making a note of this, in case I want to handle this differently
+                    # later.
+                    if len(im_out['detections']) > 0:
+                        pass
+
+                else:
+
+                    assert not class_to_assign.endswith('blank')
+
+                    # This is a scenario that's not captured well by the MD format: no detections present,
+                    # but a non-blank prediction.  For now, create a fake detection to handle this prediction.
+                    if len(im_out['detections']) == 0:
+
+                        if verbose:
+                            print('Warning: creating fake detection for non-blank whole-image classification' + \
+                                  ' in {}'.format(im_in['file']))
+                        det_out = {}
+                        all_unknown_detections.append(det_out)
+
+                        # We will change this to a string-int later
+                        det_out['category'] = 'unknown'
+                        det_out['conf'] = class_confidence
+                        det_out['bbox'] = [0,0,1,1]
+                        im_out['detections'].append(det_out)
+
+                # ...if this is/isn't a blank classification
+
+                # Attach that classification to each detection
+
+                # Create a new category ID if necessary
+                if class_to_assign in classification_category_name_to_id:
+                    classification_category_id = classification_category_name_to_id[class_to_assign]
+                else:
+                    classification_category_id = str(len(classification_category_name_to_id))
+                    classification_category_name_to_id[class_to_assign] = classification_category_id
+
+                for det in im_out['detections']:
+                    det['classifications'] = []
+                    det['classifications'].append([classification_category_id,class_confidence])
+
+            # ...if we have some type of classification for this image
+
+        # ...if this is/isn't a failure
+
+        images_out.append(im_out)
+
+    # ...for each image
+
+    if base_folder is not None:
+        if base_folder_replacements == 0:
+            print('Warning: you supplied {} as the base folder, but I made zero replacements'.format(
+                base_folder))
+
+    # Fix the 'unknown' category
+    if len(all_unknown_detections) > 0:
+
+        if len(detection_category_id_to_name) == 0:
+            max_detection_category_id = -1
+        else:
+            max_detection_category_id = max([int(x) for x in detection_category_id_to_name.keys()])
+        unknown_category_id = str(max_detection_category_id + 1)
+        detection_category_id_to_name[unknown_category_id] = 'unknown'
+
+        for det in all_unknown_detections:
+            assert det['category'] == 'unknown'
+            det['category'] = unknown_category_id
+
+
+    # Sort by filename
+
+    images_out = sort_list_of_dicts_by_key(images_out,'file')
+
+    # Prepare friendly classification names
+
+    classification_category_descriptions = \
+        invert_dictionary(classification_category_name_to_id)
+    classification_categories_out = {}
+    for category_id in classification_category_descriptions.keys():
+        category_name = classification_category_descriptions[category_id].split(';')[-1]
+        classification_categories_out[category_id] = category_name
+
+    # Prepare the output dict
+
+    detection_categories_out = detection_category_id_to_name
+    info = {}
+    info['format_version'] = 1.4
+    info['detector'] = 'converted_from_predictions_json'
+
+    if convert_human_to_person:
+        for k in detection_categories_out.keys():
+            if detection_categories_out[k] == 'human':
+                detection_categories_out[k] = 'person'
+
+    output_dict = {}
+    output_dict['info'] = info
+    output_dict['detection_categories'] = detection_categories_out
+    output_dict['classification_categories'] = classification_categories_out
+    output_dict['classification_category_descriptions'] = classification_category_descriptions
+    output_dict['images'] = images_out
+
+    if md_results_file is not None:
+        with open(md_results_file,'w') as f:
+            json.dump(output_dict,f,indent=1)
+
+        validation_options = ValidateBatchResultsOptions()
+        validation_options.raise_errors = True
+        _ = validate_batch_results(md_results_file, options=validation_options)
+
+    return output_dict
+
+# ...def generate_md_results_from_predictions_json(...)
+
+
+def generate_predictions_json_from_md_results(md_results_file,
+                                              predictions_json_file,
+                                              base_folder=None):
+    """
+    Generate a predictions.json file from the MD-formatted .json file [md_results_file].  Typically,
+    MD results files use relative paths, and predictions.json files use absolute paths, so
+    this function optionally prepends [base_folder].  Does not handle classification results in
+    MD format, since this is intended to prepare data for passing through the WI classifier.
+
+    md_to_wi.py is a command-line driver for this function.
+
+    Args:
+        md_results_file (str): path to an MD-formatted .json file
+        predictions_json_file (str): path to which we should write a predictions.json file
+        base_folder (str, optional): folder name to prepend to each path in md_results_file,
+            to convert relative paths to absolute paths.
+    """
+
+    # Validate the input file
+    validation_options = ValidateBatchResultsOptions()
+    validation_options.raise_errors = True
+    validation_options.return_data = True
+    md_results = validate_batch_results(md_results_file, options=validation_options)
+    category_id_to_name = md_results['detection_categories']
+
+    output_dict = {}
+    output_dict['predictions'] = []
+
+    # im = md_results['images'][0]
+    for im in md_results['images']:
+
+        prediction = {}
+        fn = im['file']
+        if base_folder is not None:
+            fn = os.path.join(base_folder,fn)
+        fn = fn.replace('\\','/')
+        prediction['filepath'] = fn
+        if 'failure' in im and im['failure'] is not None:
+            prediction['failures'] = ['DETECTOR']
+        else:
+            assert 'detections' in im and im['detections'] is not None
+            detections = []
+            for det in im['detections']:
+                output_det = deepcopy(det)
+                output_det['label'] = category_id_to_name[det['category']]
+                detections.append(output_det)
+
+            # detections *must* be sorted in descending order by confidence
+            detections = sort_list_of_dicts_by_key(detections,'conf', reverse=True)
+            prediction['detections'] = detections
+
+        assert len(prediction.keys()) >= 2
+        output_dict['predictions'].append(prediction)
+
+    # ...for each image
+
+    output_dir = os.path.dirname(predictions_json_file)
+    if len(output_dir) > 0:
+        os.makedirs(output_dir,exist_ok=True)
+    with open(predictions_json_file,'w') as f:
+        json.dump(output_dict,f,indent=1)
+
+# ...def generate_predictions_json_from_md_results(...)
+
+
+def generate_instances_json_from_folder(folder,
+                                        country=None,
+                                        admin1_region=None,
+                                        lat=None,
+                                        lon=None,
+                                        output_file=None,
+                                        filename_replacements=None,
+                                        tokens_to_ignore=default_tokens_to_ignore):
+    """
+    Generate an instances.json record that contains all images in [folder], optionally
+    including location information, in a format suitable for run_model.py.  Optionally writes
+    the results to [output_file].
+
+    Args:
+        folder (str): the folder to recursively search for images
+        country (str, optional): a three-letter country code
+        admin1_region (str, optional): an administrative region code, typically a two-letter
+            US state code
+        lat (float, optional): latitude to associate with all images
+        lon (float, optional): longitude to associate with all images
+        output_file (str, optional): .json file to which we should write instance records
+        filename_replacements (dict, optional): str --> str dict indicating filename substrings
+            that should be replaced with other strings.  Replacement occurs *after* converting
+            backslashes to forward slashes.
+        tokens_to_ignore (list, optional): ignore any images with these tokens in their
+            names, typically used to avoid $RECYCLE.BIN.  Can be None.
+
+    Returns:
+        dict: dict with at least the field "instances"
+    """
+
+    assert os.path.isdir(folder)
+
+    print('Enumerating images in {}'.format(folder))
+    image_files_abs = find_images(folder,recursive=True,return_relative_paths=False)
+
+    if tokens_to_ignore is not None:
+        n_images_before_ignore_tokens = len(image_files_abs)
+        for token in tokens_to_ignore:
+            image_files_abs = [fn for fn in image_files_abs if token not in fn]
+        print('After ignoring {} tokens, kept {} of {} images'.format(
+            len(tokens_to_ignore),len(image_files_abs),n_images_before_ignore_tokens))
+
+    instances = []
+
+    # image_fn_abs = image_files_abs[0]
+    for image_fn_abs in image_files_abs:
+        instance = {}
+        instance['filepath'] = image_fn_abs.replace('\\','/')
+        if filename_replacements is not None:
+            for s in filename_replacements:
+                instance['filepath'] = instance['filepath'].replace(s,filename_replacements[s])
+        if country is not None:
+            instance['country'] = country
+        if admin1_region is not None:
+            instance['admin1_region'] = admin1_region
+        if lat is not None:
+            assert lon is not None, 'Latitude provided without longitude'
+            instance['latitude'] = lat
+        if lon is not None:
+            assert lat is not None, 'Longitude provided without latitude'
+            instance['longitude'] = lon
+        instances.append(instance)
+
+    to_return = {'instances':instances}
+
+    if output_file is not None:
+        output_dir = os.path.dirname(output_file)
+        if len(output_dir) > 0:
+            os.makedirs(output_dir,exist_ok=True)
+        with open(output_file,'w') as f:
+            json.dump(to_return,f,indent=1)
+
+    return to_return
+
+# ...def generate_instances_json_from_folder(...)
+
+
+def split_instances_into_n_batches(instances_json,n_batches,output_files=None):
+    """
+    Given an instances.json file, split it into batches of equal size.
+
+    Args:
+        instances_json (str): input .json file in
+        n_batches (int): number of new files to generate
+        output_files (list, optional): output .json files for each
+            batch.  If supplied, should have length [n_batches].  If not
+            supplied, filenames will be generated based on [instances_json].
+
+    Returns:
+        list: list of output files that were written; identical to [output_files]
+        if it was supplied as input.
+    """
+
+    with open(instances_json,'r') as f:
+        instances = json.load(f)
+    assert isinstance(instances,dict) and 'instances' in instances
+    instances = instances['instances']
+
+    if output_files is not None:
+        assert len(output_files) == n_batches, \
+            'Expected {} output files, received {}'.format(
+                n_batches,len(output_files))
+    else:
+        output_files = []
+        for i_batch in range(0,n_batches):
+            batch_string = 'batch_{}'.format(str(i_batch).zfill(3))
+            output_files.append(insert_before_extension(instances_json,batch_string))
+
+    batches = split_list_into_n_chunks(instances, n_batches)
+
+    for i_batch,batch in enumerate(batches):
+        batch_dict = {'instances':batch}
+        with open(output_files[i_batch],'w') as f:
+            json.dump(batch_dict,f,indent=1)
+
+    print('Wrote {} batches to file'.format(n_batches))
+
+    return output_files
+
+# ...def split_instances_into_n_batches(...)
+
+
+def merge_prediction_json_files(input_prediction_files,output_prediction_file):
+    """
+    Merge all predictions.json files in [files] into a single .json file.
+
+    Args:
+        input_prediction_files (list): list of predictions.json files to merge
+        output_prediction_file (str): output .json file
+    """
+
+    predictions = []
+    image_filenames_processed = set()
+
+    # input_json_fn = input_prediction_files[0]
+    for input_json_fn in tqdm(input_prediction_files):
+
+        assert os.path.isfile(input_json_fn), \
+            'Could not find prediction file {}'.format(input_json_fn)
+        with open(input_json_fn,'r') as f:
+            results_this_file = json.load(f)
+        assert isinstance(results_this_file,dict)
+        predictions_this_file = results_this_file['predictions']
+        for prediction in predictions_this_file:
+            image_fn = prediction['filepath']
+            assert image_fn not in image_filenames_processed
+        predictions.extend(predictions_this_file)
+
+    output_dict = {'predictions':predictions}
+
+    output_dir = os.path.dirname(output_prediction_file)
+    if len(output_dir) > 0:
+        os.makedirs(output_dir,exist_ok=True)
+    with open(output_prediction_file,'w') as f:
+        json.dump(output_dict,f,indent=1)
+
+# ...def merge_prediction_json_files(...)
+
+
+def load_md_or_speciesnet_file(fn,verbose=True):
+    """
+    Load a .json file that may be in MD or SpeciesNet format.  Typically used so
+    SpeciesNet files can be supplied to functions originally written to support MD
+    format.
+
+    Args:
+        fn (str): a .json file in predictions.json (MD or SpeciesNet) format
+        verbose (bool, optional): enable additional debug output
+
+    Returns:
+        dict: the contents of [fn], in MD format.
+    """
+
+    with open(fn,'r') as f:
+        detector_output = json.load(f)
+
+    # If this is a SpeciesNet file, convert to MD format
+    if 'predictions' in detector_output:
+
+        if verbose:
+            print('This appears to be a SpeciesNet output file, converting to MD format')
+        detector_output = generate_md_results_from_predictions_json(predictions_json_file=fn,
+                                                                    md_results_file=None,
+                                                                    base_folder=None)
+
+    # ...if this is a SpeciesNet file
+
+    assert 'images' in detector_output, \
+        'Detector output file should be a json file with an "images" field.'
+
+    return detector_output
+
+# ...def load_md_or_speciesnet_file(...)
+
+
+def validate_predictions_file(fn,instances=None,verbose=True):
+    """
+    Validate the predictions.json file [fn].
+
+    Args:
+        fn (str): a .json file in predictions.json (SpeciesNet) format
+        instances (str or list, optional): a folder, instances.json file,
+            or dict loaded from an instances.json file.  If supplied, this
+            function will verify that [fn] contains the same number of
+            images as [instances].
+        verbose (bool, optional): enable additional debug output
+
+    Returns:
+        dict: the contents of [fn]
+    """
+
+    with open(fn,'r') as f:
+        d = json.load(f)
+    predictions = d['predictions']
+
+    failures = []
+
+    for im in predictions:
+        if 'failures' in im:
+            failures.append(im)
+
+    if verbose:
+        print('Read predictions for {} images, with {} failure(s)'.format(
+            len(d['predictions']),len(failures)))
+
+    if instances is not None:
+
+        if isinstance(instances,str):
+            if os.path.isdir(instances):
+                instances = generate_instances_json_from_folder(folder=instances)
+            elif os.path.isfile(instances):
+                with open(instances,'r') as f:
+                    instances = json.load(f)
+            else:
+                raise ValueError('Could not find instances file/folder {}'.format(
+                    instances))
+        assert isinstance(instances,dict)
+        assert 'instances' in instances
+        instances = instances['instances']
+        if verbose:
+            print('Expected results for {} files'.format(len(instances)))
+        assert len(instances) == len(predictions), \
+            '{} instances expected, {} found'.format(
+                len(instances),len(predictions))
+
+        expected_files = set([instance['filepath'] for instance in instances])
+        found_files = set([prediction['filepath'] for prediction in predictions])
+        assert expected_files == found_files
+
+    # ...if a list of instances was supplied
+
+    return d
+
+# ...def validate_predictions_file(...)
+
+
+#%% Functions related to geofencing
+
+def find_geofence_adjustments(ensemble_json_file,use_latin_names=False):
+    """
+    Count the number of instances of each unique change made by the geofence.
+
+    Args:
+        ensemble_json_file (str): SpeciesNet-formatted .json file produced
+            by the full ensemble.
+        use_latin_names (bool, optional): return a mapping using binomial names
+            rather than common names.
+
+    Returns:
+        dict: maps strings that look like "puma,felidae family" to integers,
+            where that entry would indicate the number of times that "puma" was
+            predicted, but mapped to family level by the geofence.  Sorted in
+            descending order by count.
+    """
+
+    # Load and validate ensemble results
+    ensemble_results = validate_predictions_file(ensemble_json_file)
+
+    assert isinstance(ensemble_results,dict)
+    predictions = ensemble_results['predictions']
+
+    # Maps comma-separated pairs of common names (or binomial names) to
+    # the number of times that transition (first --> second) happened
+    rollup_pair_to_count = defaultdict(int)
+
+    # prediction = predictions[0]
+    for prediction in tqdm(predictions):
+
+        if 'failures' in prediction and \
+            prediction['failures'] is not None and \
+            len(prediction['failures']) > 0:
+                continue
+
+        assert 'prediction_source' in prediction, \
+            'Prediction present without [prediction_source] field, are you sure this ' + \
+            'is an ensemble output file?'
+
+        if 'geofence' in prediction['prediction_source']:
+
+            classification_taxonomy_string = \
+                prediction['classifications']['classes'][0]
+            prediction_taxonomy_string = prediction['prediction']
+            assert is_valid_prediction_string(classification_taxonomy_string)
+            assert is_valid_prediction_string(prediction_taxonomy_string)
+
+            # Typical examples:
+            # '86f5b978-4f30-40cc-bd08-be9e3fba27a0;mammalia;rodentia;sciuridae;sciurus;carolinensis;eastern gray squirrel'
+            # 'e4d1e892-0e4b-475a-a8ac-b5c3502e0d55;mammalia;rodentia;sciuridae;;;sciuridae family'
+            classification_common_name = classification_taxonomy_string.split(';')[-1]
+            prediction_common_name = prediction_taxonomy_string.split(';')[-1]
+            classification_binomial_name = classification_taxonomy_string.split(';')[-2]
+            prediction_binomial_name = prediction_taxonomy_string.split(';')[-2]
+
+            input_name = classification_binomial_name if use_latin_names else \
+                classification_common_name
+            output_name = prediction_binomial_name if use_latin_names else \
+                prediction_common_name
+
+            rollup_pair = input_name.strip() + ',' + output_name.strip()
+            rollup_pair_to_count[rollup_pair] += 1
+
+        # ...if we made a geofencing change
+
+    # ...for each prediction
+
+    rollup_pair_to_count = sort_dictionary_by_value(rollup_pair_to_count,reverse=True)
+
+    return rollup_pair_to_count
+
+# ...def find_geofence_adjustments(...)
+
+
+def generate_geofence_adjustment_html_summary(rollup_pair_to_count,min_count=10):
+    """
+    Given a list of geofence rollups, likely generated by find_geofence_adjustments,
+    generate an HTML summary of the changes made by geofencing.  The resulting HTML
+    is wrapped in <div>, but not, for example, in <html> or <body>.
+
+    Args:
+        rollup_pair_to_count (dict): list of changes made by geofencing, see
+            find_geofence_adjustments for details
+        min_count (int, optional): minimum number of changes a pair needs in order
+            to be included in the report.
+    """
+
+    geofence_footer = ''
+
+    # Restrict to the list of taxa that were impacted by geofencing
+    rollup_pair_to_count = \
+        {key: value for key, value in rollup_pair_to_count.items() if value >= min_count}
+
+    # rollup_pair_to_count is sorted in descending order by count
+    assert is_list_sorted(list(rollup_pair_to_count.values()),reverse=True)
+
+    if len(rollup_pair_to_count) > 0:
+
+        geofence_footer = \
+            '<h3>Geofence changes that occurred more than {} times</h3>\n'.format(min_count)
+        geofence_footer += '<div class="contentdiv">\n'
+
+        print('\nRollup changes with count > {}:'.format(min_count))
+        for rollup_pair in rollup_pair_to_count.keys():
+            count = rollup_pair_to_count[rollup_pair]
+            rollup_pair_s = rollup_pair.replace(',',' --> ')
+            print('{}: {}'.format(rollup_pair_s,count))
+            rollup_pair_html = rollup_pair.replace(',',' &rarr; ')
+            geofence_footer += '{} ({})<br/>\n'.format(rollup_pair_html,count)
+
+        geofence_footer += '</div>\n'
+
+    return geofence_footer
+
+# ...def generate_geofence_adjustment_html_summary(...)
+
+
+#%% TaxonomyHandler class
+
+class TaxonomyHandler:
+    """
+    Handler for taxonomy mapping and geofencing operations.
+    """
+
+    def __init__(self, taxonomy_file, geofencing_file, country_code_file):
+        """
+        Initialize TaxonomyHandler with taxonomy information.
+
+        Args:
+            taxonomy_file (str): .csv file containing the SpeciesNet (or WI) taxonomy,
+                as seven-token taxonomic specifiers.  Distributed with the SpeciesNet model.
+            geofencing_file (str): .json file containing the SpeciesNet geofencing rules.
+                Distributed with the SpeciesNet model.
+            country_code_file: .csv file mapping country codes to names.  Should include columns
+                called "name" and "alpha-3".  A compatible file is available at
+                https://github.com/lukes/ISO-3166-Countries-with-Regional-Codes
+        """
+
+        #: Maps a taxonomy string (e.g. mammalia;cetartiodactyla;cervidae;odocoileus;virginianus) to
+        #: a dict with keys taxon_id, common_name, kingdom, phylum, class, order, family, genus, species
+        self.taxonomy_string_to_taxonomy_info = None
+
+        #: Maps a binomial name (one, two, or three ws-delimited tokens) to the same dict described above.
+        self.binomial_name_to_taxonomy_info = None
+
+        #: Maps a common name to the same dict described above
+        self.common_name_to_taxonomy_info = None
+
+        #: Dict mapping 5-token semicolon-delimited taxonomy strings to geofencing rules
+        self.taxonomy_string_to_geofencing_rules = None
+
+        #: Maps lower-case country names to upper-case country codes
+        self.country_to_country_code = None
+
+        #: Maps upper-case country codes to lower-case country names
+        self.country_code_to_country = None
+
+        self._load_taxonomy_info(taxonomy_file=taxonomy_file)
+        self._initialize_geofencing(geofencing_file=geofencing_file,
+                                    country_code_file=country_code_file)
+
+
+    def _load_taxonomy_info(self, taxonomy_file):
+        """
+        Load WI/SpeciesNet taxonomy information from a .csv file.  Stores information in the
+        instance dicts [taxonomy_string_to_taxonomy_info], [binomial_name_to_taxonomy_info],
+        and [common_name_to_taxonomy_info].
+
+        Args:
+            taxonomy_file (str): .csv file containing the SpeciesNet (or WI) taxonomy,
+                as seven-token taxonomic specifiers.  Distributed with the SpeciesNet model.
+        """
+
+        """
+        Taxonomy keys are five-token taxonomy strings, e.g.:
+
+        'mammalia;cetartiodactyla;cervidae;odocoileus;virginianus'
+
+        Taxonomy values are seven-token strings w/Taxon IDs and common names, e.g.:
+
+        '5c7ce479-8a45-40b3-ae21-7c97dfae22f5;mammalia;cetartiodactyla;cervidae;odocoileus;virginianus;white-tailed deer'
+        """
+
+        with open(taxonomy_file,'r') as f:
+            taxonomy_lines = f.readlines()
+        taxonomy_lines = [s.strip() for s in taxonomy_lines]
+
+        self.taxonomy_string_to_taxonomy_info = {}
+        self.binomial_name_to_taxonomy_info = {}
+        self.common_name_to_taxonomy_info = {}
+
+        five_token_string_to_seven_token_string = {}
+
+        for line in taxonomy_lines:
+            tokens = line.split(';')
+            assert len(tokens) == 7, 'Illegal line {} in taxonomy file {}'.format(
+                line,taxonomy_file)
+            five_token_string = ';'.join(tokens[1:-1])
+            assert len(five_token_string.split(';')) == 5
+            five_token_string_to_seven_token_string[five_token_string] = line
+
+        for taxonomy_string in five_token_string_to_seven_token_string.keys():
+
+            taxonomy_string = taxonomy_string.lower()
+
+            taxon_info = {}
+            extended_string = five_token_string_to_seven_token_string[taxonomy_string]
+            tokens = extended_string.split(';')
+            assert len(tokens) == 7
+            taxon_info['taxon_id'] = tokens[0]
+            assert len(taxon_info['taxon_id']) == 36
+            taxon_info['kingdom'] = 'animal'
+            taxon_info['phylum'] = 'chordata'
+            taxon_info['class'] = tokens[1]
+            taxon_info['order'] = tokens[2]
+            taxon_info['family'] = tokens[3]
+            taxon_info['genus'] = tokens[4]
+            taxon_info['species'] = tokens[5]
+            taxon_info['common_name'] = tokens[6]
+
+            if taxon_info['common_name'] != '':
+                self.common_name_to_taxonomy_info[taxon_info['common_name']] = taxon_info
+
+            self.taxonomy_string_to_taxonomy_info[taxonomy_string] = taxon_info
+
+            binomial_name = None
+            if len(tokens[4]) > 0 and len(tokens[5]) > 0:
+                # strip(), but don't remove spaces from the species name;
+                # subspecies are separated with a space, e.g. canis;lupus dingo
+                binomial_name = tokens[4].strip() + ' ' + tokens[5].strip()
+            elif len(tokens[4]) > 0:
+                binomial_name = tokens[4].strip()
+            elif len(tokens[3]) > 0:
+                binomial_name = tokens[3].strip()
+            elif len(tokens[2]) > 0:
+                binomial_name = tokens[2].strip()
+            elif len(tokens[1]) > 0:
+                binomial_name = tokens[1].strip()
+            if binomial_name is None:
+                # print('Warning: no binomial name for {}'.format(taxonomy_string))
+                pass
+            else:
+                self.binomial_name_to_taxonomy_info[binomial_name] = taxon_info
+
+            taxon_info['binomial_name'] = binomial_name
+
+        # ...for each taxonomy string in the file
+
+        print('Created {} records in taxonomy_string_to_taxonomy_info'.format(len(self.taxonomy_string_to_taxonomy_info)))
+        print('Created {} records in common_name_to_taxonomy_info'.format(len(self.common_name_to_taxonomy_info)))
+
+    # ...def _load_taxonomy_info(...)
+
+
+    def _initialize_geofencing(self, geofencing_file, country_code_file):
+        """
+        Load geofencing information from a .json file, and country code mappings from
+        a .csv file.  Stores results in the instance tables [taxonomy_string_to_geofencing_rules],
+        [country_to_country_code], and [country_code_to_country].
+
+        Args:
+            geofencing_file (str): .json file with geofencing rules
+            country_code_file (str): .csv file with country code mappings, in columns
+                called "name" and "alpha-3", e.g. from
+                https://github.com/lukes/ISO-3166-Countries-with-Regional-Codes/blob/master/all/all.csv
+        """
+
+        # Read country code information
+        country_code_df = pd.read_csv(country_code_file)
+        self.country_to_country_code = {}
+        self.country_code_to_country = {}
+        for i_row,row in country_code_df.iterrows():
+            self.country_to_country_code[row['name'].lower()] = row['alpha-3'].upper()
+            self.country_code_to_country[row['alpha-3'].upper()] = row['name'].lower()
+
+        # Read geofencing information
+        with open(geofencing_file,'r',encoding='utf-8') as f:
+            self.taxonomy_string_to_geofencing_rules = json.load(f)
+
+        """
+        Geofencing keys are five-token taxonomy strings, e.g.:
+
+        'mammalia;cetartiodactyla;cervidae;odocoileus;virginianus'
+
+        Geofencing values are tables mapping allow/block to country codes, optionally including region/state codes, e.g.:
+
+        {'allow': {
+          'ALA': [],
+          'ARG': [],
+          ...
+          'SUR': [],
+          'TTO': [],
+          'USA': ['AL',
+           'AR',
+           'AZ',
+           ...
+        }
+        """
+
+        # Validate
+
+        # species_string = next(iter(taxonomy_string_to_geofencing_rules.keys()))
+        for species_string in self.taxonomy_string_to_geofencing_rules.keys():
+
+            species_rules = self.taxonomy_string_to_geofencing_rules[species_string]
+
+            if len(species_rules.keys()) > 1:
+                print('Warning: taxon {} has both allow and block rules'.format(species_string))
+
+            for rule_type in species_rules.keys():
+
+                assert rule_type in ('allow','block')
+                all_country_rules_this_species = species_rules[rule_type]
+
+                for country_code in all_country_rules_this_species.keys():
+
+                    assert country_code in self.country_code_to_country
+                    region_rules = all_country_rules_this_species[country_code]
+                    # Right now we only have regional rules for the USA; these may be part of
+                    # allow or block rules.
+                    if len(region_rules) > 0:
+                        assert country_code == 'USA'
+
+                # ...for each country code in this rule set
+
+            # ...for each rule set for this species
+
+        # ...for each species
+
+    # ...def _initialize_geofencing(...)
+
+
+    def _parse_region_code_list(self, codes):
+        """
+        Turn a list of country or state codes in string, delimited string, or list format
+        into a list.  Also does basic validity checking.
+        """
+
+        if not isinstance(codes,list):
+
+            assert isinstance(codes,str)
+
+            codes = codes.strip()
+
+            # This is just a single codes
+            if ',' not in codes:
+                codes = [codes]
+            else:
+                codes = codes.split(',')
+            codes = [c.strip() for c in codes]
+
+        assert isinstance(codes,list)
+
+        codes = [c.upper().strip() for c in codes]
+
+        for c in codes:
+            assert len(c) in (2,3)
+
+        return codes
+
+    # ...def _parse_region_code_list(...)
+
+
+    def generate_csv_rows_to_block_all_countries_except(self, species_string, block_except_list):
+        """
+        Generate rows in the format expected by geofence_fixes.csv, representing a list of
+        allow and block rules to block all countries currently allowed for this species
+        except [allow_countries], and add allow rules these countries.
+
+        Args:
+            species_string (str): five-token taxonomy string
+            block_except_list (list): list of country codes not to block
+
+        Returns:
+            list of str: strings compatible with geofence_fixes.csv
+        """
+
+        assert is_valid_taxonomy_string(species_string), \
+            '{} is not a valid taxonomy string'.format(species_string)
+
+        assert self.taxonomy_string_to_geofencing_rules is not None, \
+            'Initialize geofencing prior to species lookup'
+        assert self.taxonomy_string_to_taxonomy_info is not None, \
+            'Initialize taxonomy lookup prior to species lookup'
+
+        geofencing_rules_this_species = \
+            self.taxonomy_string_to_geofencing_rules[species_string]
+
+        allowed_countries = []
+        if 'allow' in geofencing_rules_this_species:
+            allowed_countries.extend(geofencing_rules_this_species['allow'])
+
+        blocked_countries = []
+        if 'block' in geofencing_rules_this_species:
+            blocked_countries.extend(geofencing_rules_this_species['block'])
+
+        block_except_list = self._parse_region_code_list(block_except_list)
+
+        countries_to_block = []
+        countries_to_allow = []
+
+        # country = allowed_countries[0]
+        for country in allowed_countries:
+            if country not in block_except_list and country not in blocked_countries:
+                countries_to_block.append(country)
+
+        for country in block_except_list:
+            if country in blocked_countries:
+                raise ValueError("I can't allow a country that has already been blocked")
+            if country not in allowed_countries:
+                countries_to_allow.append(country)
+
+        rows = self.generate_csv_rows_for_species(species_string,
+                                             allow_countries=countries_to_allow,
+                                             block_countries=countries_to_block)
+
+        return rows
+
+    # ...def generate_csv_rows_to_block_all_countries_except(...)
+
+
+    def generate_csv_rows_for_species(self, species_string,
+                                      allow_countries=None,
+                                      block_countries=None,
+                                      allow_states=None,
+                                      block_states=None):
+        """
+        Generate rows in the format expected by geofence_fixes.csv, representing a list of
+        allow and/or block rules for the specified species and countries/states.  Does not check
+        that the rules make sense; e.g. nothing will stop you in this function from both allowing
+        and blocking a country.
+
+        Args:
+            species_string (str): five-token string in semicolon-delimited WI taxonomy format
+            allow_countries (list or str, optional): three-letter country codes, list of
+                country codes, or comma-separated list of country codes to allow
+            block_countries (list or str, optional): three-letter country codes, list of
+                country codes, or comma-separated list of country codes to block
+            allow_states (list or str, optional): two-letter state codes, list of
+                state codes, or comma-separated list of state codes to allow
+            block_states (list or str, optional): two-letter state code, list of
+                state codes, or comma-separated list of state codes to block
+
+        Returns:
+            list of str: lines ready to be pasted into geofence_fixes.csv
+        """
+
+        assert is_valid_taxonomy_string(species_string), \
+            '{} is not a valid taxonomy string'.format(species_string)
+
+        lines = []
+
+        if allow_countries is not None:
+            allow_countries = self._parse_region_code_list(allow_countries)
+            for country in allow_countries:
+                lines.append(species_string + ',allow,' + country + ',')
+
+        if block_countries is not None:
+            block_countries = self._parse_region_code_list(block_countries)
+            for country in block_countries:
+                lines.append(species_string + ',block,' + country + ',')
+
+        if allow_states is not None:
+            allow_states = self._parse_region_code_list(allow_states)
+            for state in allow_states:
+                lines.append(species_string + ',allow,USA,' + state)
+
+        if block_states is not None:
+            block_states = self._parse_region_code_list(block_states)
+            for state in block_states:
+                lines.append(species_string + ',block,USA,' + state)
+
+        return lines
+
+    # ...def generate_csv_rows_for_species(...)
+
+
+    def species_string_to_canonical_species_string(self, species):
+        """
+        Convert a string that may be a 5-token species string, a binomial name,
+        or a common name into a 5-token species string, using taxonomic lookup.
+
+        Args:
+            species (str): 5-token species string, binomial name, or common name
+
+        Returns:
+            str: common name
+
+        Raises:
+            ValueError: if [species] is not in our dictionary
+        """
+
+        species = species.lower().strip()
+
+        # Turn "species" into a taxonomy string
+
+        # If this is already a taxonomy string...
+        if len(species.split(';')) == 5:
+            taxonomy_string = species
+        # If this is a common name...
+        elif species in self.common_name_to_taxonomy_info:
+            taxonomy_info = self.common_name_to_taxonomy_info[species]
+            taxonomy_string = taxonomy_info_to_taxonomy_string(taxonomy_info)
+        # If this is a binomial name...
+        elif (species in self.binomial_name_to_taxonomy_info):
+            taxonomy_info = self.binomial_name_to_taxonomy_info[species]
+            taxonomy_string = taxonomy_info_to_taxonomy_string(taxonomy_info)
+        else:
+            raise ValueError('Could not find taxonomic information for {}'.format(species))
+
+        return taxonomy_string
+
+    # ...def species_string_to_canonical_species_string(...)
+
+
+    def species_string_to_taxonomy_info(self,species):
+        """
+        Convert a string that may be a 5-token species string, a binomial name,
+        or a common name into a taxonomic info dictionary, using taxonomic lookup.
+
+        Args:
+            species (str): 5-token species string, binomial name, or common name
+
+        Returns:
+            dict: taxonomy information
+
+        Raises:
+            ValueError: if [species] is not in our dictionary
+        """
+
+        species = species.lower().strip()
+        canonical_string = self.species_string_to_canonical_species_string(species)
+        return self.taxonomy_string_to_taxonomy_info[canonical_string]
+
+
+    def species_allowed_in_country(self, species, country, state=None, return_status=False):
+        """
+        Determines whether [species] is allowed in [country], according to
+        already-initialized geofencing rules.
+
+        Args:
+            species (str): can be a common name, a binomial name, or a species string
+            country (str): country name or three-letter code
+            state (str, optional): two-letter US state code
+            return_status (bool, optional): by default, this function returns a bool;
+                if you want to know *why* [species] is allowed/not allowed, settings
+                return_status to True will return additional information.
+
+        Returns:
+            bool or str: typically returns True if [species] is allowed in [country], else
+            False.  Returns a more detailed string if return_status is set.
+        """
+
+        assert self.taxonomy_string_to_geofencing_rules is not None, \
+            'Initialize geofencing prior to species lookup'
+        assert self.taxonomy_string_to_taxonomy_info is not None, \
+            'Initialize taxonomy lookup prior to species lookup'
+
+        taxonomy_string = self.species_string_to_canonical_species_string(species)
+
+        # Normalize [state]
+
+        if state is not None:
+            state = state.upper()
+            assert len(state) == 2
+
+        # Turn "country" into a country code
+
+        if len(country) == 3:
+            assert country.upper() in self.country_code_to_country
+            country = country.upper()
+        else:
+            assert country.lower() in self.country_to_country_code
+            country = self.country_to_country_code[country.lower()]
+
+        country_code = country.upper()
+
+        # Species with no rules are allowed everywhere
+        if taxonomy_string not in self.taxonomy_string_to_geofencing_rules:
+            status = 'allow_by_default'
+            if return_status:
+                return status
+            else:
+                return True
+
+        geofencing_rules_this_species = self.taxonomy_string_to_geofencing_rules[taxonomy_string]
+        allowed_countries = []
+        blocked_countries = []
+
+        rule_types_this_species = list(geofencing_rules_this_species.keys())
+        for rule_type in rule_types_this_species:
+            assert rule_type in ('allow','block')
+
+        if 'block' in rule_types_this_species:
+            blocked_countries = list(geofencing_rules_this_species['block'])
+        if 'allow' in rule_types_this_species:
+            allowed_countries = list(geofencing_rules_this_species['allow'])
+
+        status = None
+
+        # The convention is that block rules win over allow rules
+        if country_code in blocked_countries:
+            if country_code in allowed_countries:
+                status = 'blocked_over_allow'
+            else:
+                status = 'blocked'
+        elif country_code in allowed_countries:
+            status = 'allowed'
+        elif len(allowed_countries) > 0:
+            # The convention is that if allow rules exist, any country not on that list
+            # is blocked.
+            status = 'block_not_on_country_allow_list'
+        else:
+            # Only block rules exist for this species, and they don't include this country
+            assert len(blocked_countries) > 0
+            status = 'allow_not_on_block_list'
+
+        # Now let's see whether we have to deal with any regional rules.
+        #
+        # Right now regional rules only exist for the US.
+        if (country_code == 'USA') and ('USA' in geofencing_rules_this_species[rule_type]):
+
+            if state is None:
+
+                state_list = geofencing_rules_this_species[rule_type][country_code]
+                if len(state_list) > 0:
+                    assert status.startswith('allow')
+                    status = 'allow_no_state'
+
+            else:
+
+                state_list = geofencing_rules_this_species[rule_type][country_code]
+
+                if state in state_list:
+                    # If the state is on the list, do what the list says
+                    if rule_type == 'allow':
+                        status = 'allow_on_state_allow_list'
+                    else:
+                        status = 'block_on_state_block_list'
+                else:
+                    # If the state is not on the list, do the opposite of what the list says
+                    if rule_type == 'allow':
+                        status = 'block_not_on_state_allow_list'
+                    else:
+                        status = 'allow_not_on_state_block_list'
+
+        if return_status:
+            return status
+        else:
+            if status.startswith('allow'):
+                return True
+            else:
+                assert status.startswith('block')
+                return False
+
+    # ...def species_allowed_in_country(...)
+
+
+    def export_geofence_data_to_csv(self, csv_fn=None, include_common_names=True):
+        """
+        Converts the geofence .json representation into an equivalent .csv representation,
+        with one taxon per row and one region per column.  Empty values indicate non-allowed
+        combinations, positive numbers indicate allowed combinations.  Negative values
+        are reserved for specific non-allowed combinations.
+
+        Args:
+            csv_fn (str): output .csv file
+            include_common_names (bool, optional): include a column for common names
+
+        Returns:
+            dataframe: the pandas representation of the csv output file
+        """
+
+        all_taxa = sorted(list(self.taxonomy_string_to_geofencing_rules.keys()))
+        print('Preparing geofencing export for {} taxa'.format(len(all_taxa)))
+
+        all_regions = set()
+
+        # taxon = all_taxa[0]
+        for taxon in all_taxa:
+
+            taxon_rules = self.taxonomy_string_to_geofencing_rules[taxon]
+            for rule_type in taxon_rules.keys():
+
+                assert rule_type in ('allow','block')
+                all_country_rules_this_species = taxon_rules[rule_type]
+
+                for country_code in all_country_rules_this_species.keys():
+                    all_regions.add(country_code)
+                    assert country_code in self.country_code_to_country
+                    assert len(country_code) == 3
+                    region_rules = all_country_rules_this_species[country_code]
+                    if len(region_rules) > 0:
+                        assert country_code == 'USA'
+                        for region_name in region_rules:
+                            assert len(region_name) == 2
+                            assert isinstance(region_name,str)
+                            all_regions.add(country_code + ':' + region_name)
+
+        all_regions = sorted(list(all_regions))
+
+        print('Found {} regions'.format(len(all_regions)))
+
+        n_allowed = 0
+        df = pd.DataFrame(index=all_taxa,columns=all_regions)
+        # df = df.fillna(np.nan)
+
+        for taxon in tqdm(all_taxa):
+            for region in all_regions:
+                tokens = region.split(':')
+                country_code = tokens[0]
+                state_code = None
+                if len(tokens) > 1:
+                    state_code = tokens[1]
+                allowed = self.species_allowed_in_country(species=taxon,
+                                                     country=country_code,
+                                                     state=state_code,
+                                                     return_status=False)
+                if allowed:
+                    n_allowed += 1
+                    df.loc[taxon,region] = 1
+
+            # ...for each region
+
+        # ...for each taxon
+
+        print('Allowed {} of {} combinations'.format(n_allowed,len(all_taxa)*len(all_regions)))
+
+        # Before saving, convert columns with numeric values to integers
+        for col in df.columns:
+            # Check whether each column has any non-NaN values that could be integers
+            if df[col].notna().any() and pd.to_numeric(df[col], errors='coerce').notna().any():
+                # Convert column to Int64 type (pandas nullable integer type)
+                df[col] = pd.to_numeric(df[col], errors='coerce').astype('Int64')
+
+        if include_common_names:
+            df.insert(loc=0,column='common_name',value='')
+            for taxon in all_taxa:
+                if taxon in self.taxonomy_string_to_taxonomy_info:
+                    taxonomy_info = self.taxonomy_string_to_taxonomy_info[taxon]
+                    common_name = taxonomy_info['common_name']
+                    assert isinstance(common_name,str) and len(common_name) < 50
+                    df.loc[taxon,'common_name'] = common_name
+
+        if csv_fn is not None:
+            df.to_csv(csv_fn,index=True,header=True)
+
+        return df
+
+    # ...def export_geofence_data_to_csv(...)
+
+# ...class TaxonomyHandler