Merge pull request #600 from wright-group/development

3.0.2

.github/CODEOWNERS

@@ -1,3 +1,3 @@
 # Require review from one maintainer for pushes to Development and Master
 
-* @untzag @ksunden @p770193
+* @untzag @ksunden @p770193 @darienmorrow

CONTRIBUTORS

@@ -1,10 +1,11 @@
 Maintainers (alphabetical by last name)
   Daniel Kohler (@p770193)
+  Darien Morrow <darienmorrow@gmail.com> (@darienmorrow)
   Kyle Sunden <sunden@wisc.edu> (@ksunden)
   Blaise Thompson <blaise@untzag.com> (@untzag)
 
 Contributors (alphabetical by last name)
   Kyle Czech (@kjczech)
-  Darien Morrow <darienmorrow@gmail.com> (@darienmorrow)
   Nathan Neff-Mallon (@neffmallon)
+  Thomas Parker (@parkertomf)
   Rachel Swedin (@RachelSwedin)

WrightTools/VERSION

@@ -0,0 +1 @@
+3.0.2

WrightTools/__version__.py

@@ -20,7 +20,7 @@
 
 
 # read from VERSION file
-with open(os.path.join(os.path.dirname(here), 'VERSION')) as f:
+with open(os.path.join(here, 'VERSION')) as f:
     __version__ = f.read().strip()
 
 
@@ -30,6 +30,6 @@
     with open(p) as f:
         __branch__ = f.readline().rstrip().split(r'/')[-1]
     if __branch__ != 'master':
-        __version__ += '-' + __branch__
+        __version__ += '+' + __branch__
 else:
     __branch__ = None

WrightTools/__wt5_version__.py

@@ -20,5 +20,5 @@
 
 
 # read from VERSION file
-with open(os.path.join(os.path.dirname(here), 'WT5_VERSION')) as f:
+with open(os.path.join(here, 'WT5_VERSION')) as f:
     __wt5_version__ = f.read().strip()

WrightTools/collection/_cary.py

@@ -70,8 +70,16 @@ def from_Cary(filepath, name=None, parent=None, verbose=True):
             if line == '\n' or line == '':
                 break
             else:
+                # Note, it is necessary to call this twice, as a single call will 
+                # result in something like ',,,,' > ',nan,,nan,'.
+                line = line.replace(',,', ',nan,')
+                line = line.replace(',,', ',nan,')
+                # Ensure that the first column has nan, if necessary
+                if line[0] == ',':
+                    line = 'nan' + line
                 clean = line[:-2]  # lines end with ',/n'
                 lines.append(np.fromstring(clean, sep=','))
+    lines = [line for line in lines if len(line) > 0]
     header = header.split(',')
     columns = columns.split(',')
     arr = np.array(lines).T
@@ -83,8 +91,9 @@ def from_Cary(filepath, name=None, parent=None, verbose=True):
         ax = spl[0].lower() if len(spl) > 0 else None
         units = spl[1].lower() if len(spl) > 1 else None
         dat = datas.create_data(header[i], kind='Cary', source=filepath)
-        dat.create_variable(ax, arr[i], units=units)
-        dat.create_channel(columns[i + 1].lower(), arr[i + 1], label=columns[i + 1].lower())
+        dat.create_variable(ax, arr[i][~np.isnan(arr[i])], units=units)
+        dat.create_channel(columns[i + 1].lower(), arr[i + 1][~np.isnan(arr[i + 1])],
+                           label=columns[i + 1].lower())
         dat.transform(ax)
     # finish
     if verbose:

WrightTools/data/_join.py

@@ -52,23 +52,29 @@ def join(datas, *, name='join',  parent=None, verbose=True):
     datas = list(datas)
     # check if variables are valid
     axis_expressions = datas[0].axis_expressions
+    variable_names = set(datas[0].variable_names)
+    channel_names = set(datas[0].channel_names)
+    for d in datas:
+        variable_names &= set(d.variable_names)
+        channel_names &= set(d.channel_names)
+    variable_names = list(variable_names)
+    channel_names = list(channel_names)
     variable_units = []
-    variable_names = []
-    for a in datas[0].axes:
-        for v in a.variables:
-            variable_names.append(v.natural_name)
-            variable_units.append(v.units)
-    # TODO: check if all other datas have the same variable names
-    # check if channels are valid
-    # TODO: this is a hack
     channel_units = []
-    channel_names = []
-    for c in datas[0].channels:
-        channel_names.append(c.natural_name)
-        channel_units.append(c.units)
+    for v in variable_names:
+        variable_units.append(datas[0][v].units)
+    for c in channel_names:
+        channel_units.append(datas[0][v].units)
     # axis variables
+    axis_variable_names = []
+    axis_variable_units = []
+    for a in datas[0].axes:
+        for v in a.variables:
+            axis_variable_names.append(v.natural_name)
+            axis_variable_units.append(v.units)
+
     vs = collections.OrderedDict()
-    for n, units in zip(variable_names, variable_units):
+    for n, units in zip(axis_variable_names, axis_variable_units):
         values = np.concatenate([d[n][:].flat for d in datas])
         rounded = values.round(8)
         _, idxs = np.unique(rounded, True)
@@ -90,11 +96,13 @@ def from_dict(d, parent=None):
             var = out.create_variable(values=values, **datas[0][k].attrs)
             i += 1
         return out
+
+
     out = from_dict(vs, parent=parent)
     for channel_name, units in zip(channel_names, channel_units):
         # **attrs passes the name and units as well
         out.create_channel(**datas[0][channel_name].attrs)
-    for variable_name in datas[0].variable_names:
+    for variable_name in variable_names:
         if variable_name not in vs.keys():
             shape = tuple(1 if i == 1 else n for i, n in zip(datas[0][variable_name].shape,
                                                              out.shape))

WrightTools/data/_pycmds.py

@@ -4,18 +4,12 @@
 # --- import --------------------------------------------------------------------------------------
 
 
-import collections
-import warnings
-
 import numpy as np
 
-from scipy.interpolate import griddata
-
 import tidy_headers
 
 from ._data import Data
 from .. import kit as wt_kit
-from .. import units as wt_units
 
 
 # --- define --------------------------------------------------------------------------------------
@@ -82,6 +76,16 @@ def from_PyCMDS(filepath, name=None, parent=None, verbose=True):
                 'centers': centers}
         axes.append(axis)
     shape = tuple([a['points'].size for a in axes])
+    for i, ax in enumerate(axes):
+        sh = [1] * len(shape)
+        sh[i] = len(ax['points'])
+        data.create_variable(name=ax['name'] + '_points',
+                             values=np.array(ax['points']).reshape(sh))
+        if ax['centers'] is not None:
+            sh = [1] * len(shape)
+            sh[i - 1] = len(axes[i - 1]['points'])
+            data.create_variable(name=ax['name'] + '_centers',
+                                 values=np.array(ax['centers']).reshape(sh))
     # get assorted remaining things
     # variables and channels
     for index, kind, name in zip(range(len(arr)), headers['kind'], headers['name']):
@@ -112,8 +116,13 @@ def from_PyCMDS(filepath, name=None, parent=None, verbose=True):
                         tolerance = 3.
                     if 'zero' in name:
                         tolerance = 1e-10
-                    if name in headers['axis names']:
-                        tolerance = 1e-5
+                    try:
+                        assert i == headers['axis names'].index(name)
+                        tolerance = 0
+                    except (ValueError, AssertionError):
+                        if (name in headers['axis names'] and
+                                '%s_centers' % name not in data.variable_names):
+                            tolerance = np.inf
                     mean = np.nanmean(values, axis=i)
                     mean = np.expand_dims(mean, i)
                     values, meanexp = wt_kit.share_nans(values, mean)