Merge pull request #3 from SmithB/master

Clouds_and_data_filtering added

Author: dshean

05_Clouds_and_ICESat-2_Data_Filtering/.gitignore

@@ -0,0 +1,5 @@
+*checkpoint*
+*.Copy*.*
+*_pycache_*
+.nfs*
+

05_Clouds_and_ICESat-2_Data_Filtering/notebooks/Clouds_and_data_quality.ipynb

@@ -0,0 +1,53 @@
+import numpy as np
+import h5py
+
+
+def ATL06_to_dict(filename, dataset_dict):
+    """
+        Read selected datasets from an ATL06 file
+
+        Input arguments:
+            filename: ATl06 file to read
+            dataset_dict: A dictinary describing the fields to be read
+                    keys give the group names to be read, 
+                    entries are lists of datasets within the groups
+        Output argument:
+            D6: dictionary containing ATL06 data.  Each dataset in 
+                dataset_dict has its own entry in D6.  Each dataset 
+                in D6 contains a list of numpy arrays containing the 
+                data
+    """
+    
+    D6=[]
+    pairs=[1, 2, 3]
+    beams=['l','r']
+    # open the HDF5 file
+    with h5py.File(filename) as h5f:
+        # loop over beam pairs
+        for pair in pairs:
+            # loop over beams
+            for beam_ind, beam in enumerate(beams):
+                # check if a beam exists, if not, skip it
+                if '/gt%d%s/land_ice_segments' % (pair, beam) not in h5f:
+                    continue
+                # loop over the groups in the dataset dictionary
+                temp={}
+                for group in dataset_dict.keys():
+                    for dataset in dataset_dict[group]:
+                        DS='/gt%d%s/%s/%s' % (pair, beam, group, dataset)
+                        # since a dataset may not exist in a file, we're going to try to read it, and if it doesn't work, we'll move on to the next:
+                        try:
+                            temp[dataset]=np.array(h5f[DS])
+                            # some parameters have a _FillValue attribute.  If it exists, use it to identify bad values, and set them to np.NaN
+                            if '_FillValue' in h5f[DS].attrs:
+                                fill_value=h5f[DS].attrs['_FillValue']
+                                temp[dataset][temp[dataset]==fill_value]=np.NaN
+                        except KeyError as e:
+                            pass
+                if len(temp) > 0:
+                    # it's sometimes convenient to have the beam and the pair as part of the output data structure: This is how we put them there.
+                    temp['pair']=np.zeros_like(temp['h_li'])+pair
+                    temp['beam']=np.zeros_like(temp['h_li'])+beam_ind
+                    temp['filename']=filename
+                    D6.append(temp)
+    return D6