From c014a0df4b155bff60780732040baa9ca1a851d2 Mon Sep 17 00:00:00 2001
From: Bert Palm <bert.palm@ufz.de>
Date: Thu, 12 Mar 2020 14:43:30 +0100
Subject: [PATCH] fine new properties / methods
---
dios/dios.py | 60 ++++++++++++++++++++++++++++++++++++++++++++++------
1 file changed, 54 insertions(+), 6 deletions(-)
diff --git a/dios/dios.py b/dios/dios.py
index 0675e3d..8117f3c 100644
--- a/dios/dios.py
+++ b/dios/dios.py
@@ -13,7 +13,6 @@ import numpy as np
import operator as op
import functools as ftools
-import itertools
import pandas.core.dtypes.common as dcom
@@ -127,9 +126,8 @@ class DictOfSeries:
if data is not None:
self._init_insert_data(data, columns)
- # NOTE: self._data contain nans at locations
- # where no data was present, but a column-name
- # was given
+ # self._data contain nans at locations, where no
+ # data was present, but a column-name was given
if itype is None:
self._itype = _find_least_common_itype(self._data.dropna())
@@ -189,6 +187,48 @@ class DictOfSeries:
def columns(self, cols):
self._data.index = cols
+ @property
+ def indexes(self):
+ """ Return pd.series with the indexes. """
+ return self._fast_apply('index')
+
+ def index_of(self, method='all'):
+ """ Return an single index with indices from all columns.
+
+ parameter
+ ----------
+ method: string
+ - 'all' : get all indices from all columns
+ - 'union' : alias for 'all'
+ - 'shared' : get indices that are present in every columns
+ - 'intersection' : alias for 'shared'
+ - 'uniques' : get indices that are only present in a single column
+ - 'non-uniques' : get indices that are present in more than one column
+
+ Note
+ ----
+ The returned index never contains duplicates.
+ """
+
+ indexes = self.indexes
+ if len(indexes) <= 1:
+ return indexes.squeeze()
+
+ if method in ['union', 'all']:
+ res = ftools.reduce(pd.Index.union, indexes)
+ elif method == ['intersection', 'shared']:
+ res = ftools.reduce(pd.Index.intersection, indexes)
+ elif method in ['uniques', 'non-uniques']:
+ res = ftools.reduce(pd.Index.append, indexes)
+ res = res.value_counts(sort=False, dropna=False)
+ if method == 'uniques':
+ res = res[res == 1].index
+ else:
+ res = res[res > 1].index
+ else:
+ raise ValueError(method)
+ return res if res.is_unique else res.unique()
+
@property
def itype(self):
return self._itype
@@ -352,6 +392,14 @@ class DictOfSeries:
raise ValueError(axis)
+ @property
+ def size(self):
+ return self.lengths.sum()
+
+ @property
+ def lengths(self):
+ return self._data.apply(len)
+
def __len__(self):
return len(self.columns)
@@ -486,8 +534,8 @@ class DictOfSeries:
if not isinstance(d, pd.Series):
d = pd.Series(d)
data.at[c] = d
- new = DictOfSeries(data=data, itype=MixedItype, fastpath=True)
- new._itype = new.__find_least_common_itype()
+ itype = _find_least_common_itype(data)
+ new = DictOfSeries(data=data, itype=itype, fastpath=True)
else:
new = pd.Series(data=new, index=self.columns)
else:
--
GitLab