trufl 0.0.2__py3-none-any.whl

trufl/__init__.py

@@ -0,0 +1 @@
+__version__ = "0.0.2"

trufl/_modidx.py

@@ -0,0 +1,81 @@
+# Autogenerated by nbdev
+
+d = { 'settings': { 'branch': 'main',
+                'doc_baseurl': '/trufl',
+                'doc_host': 'https://franckalbinet.github.io',
+                'git_url': 'https://github.com/franckalbinet/trufl',
+                'lib_path': 'trufl'},
+  'syms': { 'trufl.callbacks': { 'trufl.callbacks.Callback': ('callbacks.html#callback', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.CountCB': ('callbacks.html#countcb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.CountCB.__call__': ('callbacks.html#countcb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.CountCB.__init__': ('callbacks.html#countcb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MaxCB': ('callbacks.html#maxcb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MaxCB.__call__': ('callbacks.html#maxcb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MaxCB.__init__': ('callbacks.html#maxcb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MinCB': ('callbacks.html#mincb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MinCB.__call__': ('callbacks.html#mincb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MinCB.__init__': ('callbacks.html#mincb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MoranICB': ('callbacks.html#moranicb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MoranICB.__call__': ('callbacks.html#moranicb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MoranICB.__init__': ('callbacks.html#moranicb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.MoranICB._weights': ('callbacks.html#moranicb._weights', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.PriorCB': ('callbacks.html#priorcb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.PriorCB.__call__': ('callbacks.html#priorcb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.PriorCB.__init__': ('callbacks.html#priorcb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State': ('callbacks.html#state', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State.__call__': ('callbacks.html#state.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State.__init__': ('callbacks.html#state.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State._flatten': ('callbacks.html#state._flatten', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State.expand_to_k_nearest': ( 'callbacks.html#state.expand_to_k_nearest',
+                                                                                'trufl/callbacks.py'),
+                                 'trufl.callbacks.State.get': ('callbacks.html#state.get', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.State.run_cbs': ('callbacks.html#state.run_cbs', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.StdCB': ('callbacks.html#stdcb', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.StdCB.__call__': ('callbacks.html#stdcb.__call__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.StdCB.__init__': ('callbacks.html#stdcb.__init__', 'trufl/callbacks.py'),
+                                 'trufl.callbacks.Variable': ('callbacks.html#variable', 'trufl/callbacks.py')},
+            'trufl.collector': { 'trufl.collector.DataCollector': ('collector.html#datacollector', 'trufl/collector.py'),
+                                 'trufl.collector.DataCollector.__init__': ('collector.html#datacollector.__init__', 'trufl/collector.py'),
+                                 'trufl.collector.DataCollector.collect': ('collector.html#datacollector.collect', 'trufl/collector.py'),
+                                 'trufl.collector.DataCollector.get_values': ( 'collector.html#datacollector.get_values',
+                                                                               'trufl/collector.py')},
+            'trufl.mcdm': { 'trufl.mcdm.abspearson': ('mcdm.html#abspearson', 'trufl/mcdm.py'),
+                            'trufl.mcdm.check_normalization_input': ('mcdm.html#check_normalization_input', 'trufl/mcdm.py'),
+                            'trufl.mcdm.check_scoring_input': ('mcdm.html#check_scoring_input', 'trufl/mcdm.py'),
+                            'trufl.mcdm.check_weighting_input': ('mcdm.html#check_weighting_input', 'trufl/mcdm.py'),
+                            'trufl.mcdm.correlate': ('mcdm.html#correlate', 'trufl/mcdm.py'),
+                            'trufl.mcdm.cp': ('mcdm.html#cp', 'trufl/mcdm.py'),
+                            'trufl.mcdm.critic': ('mcdm.html#critic', 'trufl/mcdm.py'),
+                            'trufl.mcdm.dcor': ('mcdm.html#dcor', 'trufl/mcdm.py'),
+                            'trufl.mcdm.dist_matrix': ('mcdm.html#dist_matrix', 'trufl/mcdm.py'),
+                            'trufl.mcdm.em': ('mcdm.html#em', 'trufl/mcdm.py'),
+                            'trufl.mcdm.is_normalized_matrix': ('mcdm.html#is_normalized_matrix', 'trufl/mcdm.py'),
+                            'trufl.mcdm.is_normalized_vector': ('mcdm.html#is_normalized_vector', 'trufl/mcdm.py'),
+                            'trufl.mcdm.lin_func': ('mcdm.html#lin_func', 'trufl/mcdm.py'),
+                            'trufl.mcdm.linear1': ('mcdm.html#linear1', 'trufl/mcdm.py'),
+                            'trufl.mcdm.linear2': ('mcdm.html#linear2', 'trufl/mcdm.py'),
+                            'trufl.mcdm.linear3': ('mcdm.html#linear3', 'trufl/mcdm.py'),
+                            'trufl.mcdm.mw': ('mcdm.html#mw', 'trufl/mcdm.py'),
+                            'trufl.mcdm.normalize': ('mcdm.html#normalize', 'trufl/mcdm.py'),
+                            'trufl.mcdm.pearson': ('mcdm.html#pearson', 'trufl/mcdm.py'),
+                            'trufl.mcdm.score': ('mcdm.html#score', 'trufl/mcdm.py'),
+                            'trufl.mcdm.sd': ('mcdm.html#sd', 'trufl/mcdm.py'),
+                            'trufl.mcdm.squared_dcor': ('mcdm.html#squared_dcor', 'trufl/mcdm.py'),
+                            'trufl.mcdm.squared_dcov': ('mcdm.html#squared_dcov', 'trufl/mcdm.py'),
+                            'trufl.mcdm.squared_dcov_matrix': ('mcdm.html#squared_dcov_matrix', 'trufl/mcdm.py'),
+                            'trufl.mcdm.topsis': ('mcdm.html#topsis', 'trufl/mcdm.py'),
+                            'trufl.mcdm.vector': ('mcdm.html#vector', 'trufl/mcdm.py'),
+                            'trufl.mcdm.vic': ('mcdm.html#vic', 'trufl/mcdm.py'),
+                            'trufl.mcdm.weigh': ('mcdm.html#weigh', 'trufl/mcdm.py')},
+            'trufl.optimizer': { 'trufl.optimizer.Optimizer': ('optimizer.html#optimizer', 'trufl/optimizer.py'),
+                                 'trufl.optimizer.Optimizer.__init__': ('optimizer.html#optimizer.__init__', 'trufl/optimizer.py'),
+                                 'trufl.optimizer.Optimizer.rank': ('optimizer.html#optimizer.rank', 'trufl/optimizer.py')},
+            'trufl.reader': { 'trufl.reader.read_geojson': ('reader.html#read_geojson', 'trufl/reader.py'),
+                              'trufl.reader.read_shapefile': ('reader.html#read_shapefile', 'trufl/reader.py')},
+            'trufl.sampler': { 'trufl.sampler.Sampler': ('sampler.html#sampler', 'trufl/sampler.py'),
+                               'trufl.sampler.Sampler.__init__': ('sampler.html#sampler.__init__', 'trufl/sampler.py'),
+                               'trufl.sampler.Sampler.loc_ids': ('sampler.html#sampler.loc_ids', 'trufl/sampler.py'),
+                               'trufl.sampler.Sampler.sample': ('sampler.html#sampler.sample', 'trufl/sampler.py')},
+            'trufl.utils': { 'trufl.utils.anonymize_raster': ('utils.html#anonymize_raster', 'trufl/utils.py'),
+                             'trufl.utils.gridder': ('utils.html#gridder', 'trufl/utils.py'),
+                             'trufl.utils.reproject_raster': ('utils.html#reproject_raster', 'trufl/utils.py')}}}

trufl/callbacks.py

@@ -0,0 +1,184 @@
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/04_callbacks.ipynb.
+
+# %% auto 0
+__all__ = ['Variable', 'Callback', 'State', 'MaxCB', 'MinCB', 'StdCB', 'CountCB', 'MoranICB', 'PriorCB']
+
+# %% ../nbs/04_callbacks.ipynb 2
+from dataclasses import dataclass
+
+import warnings
+warnings.filterwarnings("ignore", category=UserWarning, message="The weights matrix is not fully connected")
+
+# %% ../nbs/04_callbacks.ipynb 3
+from pysal.lib import weights
+from pysal.explore import esda
+import itertools
+import fastcore.all as fc
+from fastcore.basics import patch
+import numpy as np
+from scipy.spatial import KDTree
+import geopandas as gpd
+from typing import List
+from collections.abc import Callable
+import rasterio
+from rasterio.mask import mask
+import pandas as pd
+from typing import Type
+
+# %% ../nbs/04_callbacks.ipynb 5
+@dataclass
+class Variable:
+    "State variable"
+    name: str
+    value: float
+
+# %% ../nbs/04_callbacks.ipynb 6
+class Callback(): pass
+
+# %% ../nbs/04_callbacks.ipynb 8
+class State:
+    def __init__(self, 
+                 measurements:gpd.GeoDataFrame, # Measurements data with `loc_id`, `geometry` and `value` columns. 
+                 smp_areas:gpd.GeoDataFrame, # Grid of areas/polygons of interest with `loc_id` and `geometry`.
+                 cbs:List[Callable], # List of Callback functions returning `Variable`s.
+                ): 
+        "Collect various variables/metrics per grid cell/administrative unit."
+        fc.store_attr()
+        self.unsampled_locs = self.smp_areas.index.difference(self.measurements.index)
+
+# %% ../nbs/04_callbacks.ipynb 9
+@patch
+def get(self:State, 
+        loc_id:str, # Unique id of the Point feature
+        as_numpy=False # Whether or not to return a list of `Variable` or a tuple of numpy arrays.
+       ):
+    "Get the state variables as defined by `cbs` for a given location (`loc_id`)."
+    variables = self.run_cbs(loc_id)
+    if as_numpy:
+        return (np.array([v.name for v in variables]), 
+                np.array([v.value for v in variables]))
+    else:
+        return variables
+
+# %% ../nbs/04_callbacks.ipynb 10
+@patch
+def __call__(self:State, loc_id=None, **kwargs):
+    "Get the state variables as defined by `cbs` for all `loc_id`s as a dataframe."
+    loc_ids = self.smp_areas.index
+    results = [{v.name: v.value for v in self.run_cbs(loc_id)} | {'loc_id': loc_id} for loc_id in loc_ids]
+    return pd.DataFrame(results).set_index('loc_id')
+
+# %% ../nbs/04_callbacks.ipynb 11
+@patch
+def expand_to_k_nearest(self:State, 
+                        subset_measurements:gpd.GeoDataFrame, # Measurements for which Variables are computed.
+                        k:int=5, # Number of nearest neighbours (possibly belonging to adjacent cells/admin. units to consider).
+                       ):
+    "Expand measurements of concern possibly to nearest neighbors of surrounding grid cells."
+    tree = KDTree(self.measurements.geometry.apply(lambda p: (p.x, p.y)).tolist());
+    _, indices = tree.query(subset_measurements.geometry.apply(lambda p: (p.x, p.y)).tolist(), k=k)
+    return self.measurements.iloc[indices.flatten()].reset_index(drop=True)
+
+# %% ../nbs/04_callbacks.ipynb 12
+@patch
+def _flatten(self:State, variables):
+    "Flatten list of variables potentially containing both scalar and tuples."
+    return list(itertools.chain(*(v if isinstance(v, tuple) else (v,) 
+                                  for v in variables)))
+
+# %% ../nbs/04_callbacks.ipynb 13
+@patch
+def run_cbs(self:State, loc_id):
+    "Run Callbacks sequentially and flatten the results if required."
+    variables = []
+    for cb in self.cbs:
+        variables.append(cb(loc_id, self))
+    return self._flatten(variables)
+
+# %% ../nbs/04_callbacks.ipynb 15
+class MaxCB(Callback):
+    "Compute Maximum value of measurements at given location."
+    def __init__(self, name='Max'): fc.store_attr()
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        if loc_id in o.unsampled_locs: return Variable(self.name, np.nan)
+        return Variable(self.name, 
+                        np.max(o.measurements.loc[[loc_id]].value.values))
+
+# %% ../nbs/04_callbacks.ipynb 16
+class MinCB(Callback):
+    "Compute Minimum value of measurements at given location."
+    def __init__(self, name='Min'): fc.store_attr()
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        if loc_id in o.unsampled_locs: return Variable(self.name, np.nan)
+        return Variable(self.name, 
+                    np.min(o.measurements.loc[[loc_id]].value.values))
+
+# %% ../nbs/04_callbacks.ipynb 17
+class StdCB(Callback):
+    "Compute Standard deviation of measurements at given location."
+    def __init__(self, name='Standard Deviation'): fc.store_attr()
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        if loc_id in o.unsampled_locs: return Variable(self.name, np.nan)
+        return Variable(self.name, 
+                    np.std(o.measurements.loc[[loc_id]].value.values))
+
+# %% ../nbs/04_callbacks.ipynb 18
+class CountCB(Callback):
+    "Compute the number of measurements at given location."
+    def __init__(self, name='Count'): fc.store_attr()
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        if loc_id in o.unsampled_locs: return Variable(self.name, np.nan)
+        return Variable(self.name, 
+                        len(o.measurements.loc[[loc_id]].value.values))
+
+# %% ../nbs/04_callbacks.ipynb 19
+class MoranICB(Callback):
+    "Compute Moran.I of measurements at given location. Return NaN if p_sim above threshold."
+    def __init__(self, k=5, p_threshold=0.05, name='Moran.I', min_n=5): fc.store_attr()
+
+    def _weights(self, measurements):
+        w = weights.KNN.from_dataframe(measurements, k=self.k)
+        w.transform = "R" # Row-standardization
+        return w
+        
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        if loc_id in o.unsampled_locs: return Variable(self.name, np.nan)
+        subset = o.measurements.loc[[loc_id]]
+        if len(subset) <= self.min_n: return Variable(self.name, np.nan)
+        expanded_measurements = o.expand_to_k_nearest(subset, k=self.k)
+        moran = esda.moran.Moran(expanded_measurements['value'], self._weights(expanded_measurements))
+        return Variable(self.name, moran.I if moran.p_sim < self.p_threshold else np.nan)
+
+# %% ../nbs/04_callbacks.ipynb 20
+class PriorCB(Callback):
+    "Emulate a prior by taking the mean of measurement over a single grid cell."
+    def __init__(self, 
+                 fname_raster:str, # Name of raster file
+                 name:str='Prior' # Name of the State variable
+                ): 
+        fc.store_attr()
+
+    def __call__(self, 
+                 loc_id:int, # Unique id of an individual area of interest. 
+                 o:Type[State] # A State's object
+                ): 
+        polygon = o.smp_areas.loc[o.smp_areas.reset_index().loc_id == loc_id].geometry
+        with rasterio.open(self.fname_raster) as src:
+            out_image, out_transform = mask(src, polygon, crop=True)
+            mean_value = np.mean(out_image)
+        return Variable(self.name, mean_value)

trufl/collector.py

@@ -0,0 +1,35 @@
+# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/06_collector.ipynb.
+
+# %% auto 0
+__all__ = ['DataCollector']
+
+# %% ../nbs/06_collector.ipynb 3
+import rasterio
+import fastcore.all as fc
+import geopandas as gpd
+from rasterio import transform
+
+# %% ../nbs/06_collector.ipynb 5
+class DataCollector:
+    def __init__(self, 
+                 fname_raster:str, # The path to the raster file.
+                 band:int=1, # The band number to use. Defaults to 1.
+                ):
+        "Emulate data collection. Provided a set of location, return values sampled from given raster file."
+        fc.store_attr()
+        with rasterio.open(fname_raster) as src:
+            self.band_data = src.read(band)
+            self.affine = src.transform
+            self.bounds = src.bounds
+    
+    def get_values(self, 
+                   gdf:gpd.GeoDataFrame # loc_id and Point/Multipoint geometry of samples where to measure.
+                  ):
+        coords = [(x, y) for x, y in gdf.get_coordinates().values]
+        pixel_coords = [transform.rowcol(self.affine, *pair) for pair in coords]
+        return [self.band_data[int(x), int(y)] for (x, y) in pixel_coords]
+        
+    def collect(self, 
+                gdf:gpd.GeoDataFrame # loc_id and Point/Multipoint geometry of samples where to measure.
+               ) -> gpd.GeoDataFrame:
+        return gdf.explode(index_parts=False).assign(value=self.get_values(gdf))