From 20cac5df6e3ff5d43b1595ec63b57c2f459c25b3 Mon Sep 17 00:00:00 2001
From: Bert Palm <bert.palm@ufz.de>
Date: Mon, 13 Apr 2020 03:12:30 +0200
Subject: [PATCH] added index to init. make insert in dios more strict.
---
dios/dios.py | 38 +++++++++++++++++++++++++-------------
1 file changed, 25 insertions(+), 13 deletions(-)
diff --git a/dios/dios.py b/dios/dios.py
index 63c08bd..ad16abd 100644
--- a/dios/dios.py
+++ b/dios/dios.py
@@ -66,7 +66,7 @@ class DictOfSeries:
Policy to use for down-casting an itype.
"""
- def __init__(self, data=None, columns=None, itype=None, cast_policy='save', fastpath=False):
+ def __init__(self, data=None, columns=None, index=None, itype=None, cast_policy='save', fastpath=False):
# we are called internally
if fastpath:
@@ -82,22 +82,28 @@ class DictOfSeries:
self._data = pd.Series(dtype='O', index=columns)
else:
+ if index is not None and not isinstance(index, pd.Index):
+ index = pd.Index(index)
+
# itype=None means infer the itype by the data, so we first set to the highest
# possible itype, then insert data, then find the best-fitting itype.
- if itype is None:
+ if itype is None and index is None:
self._itype = ObjItype
else:
- self._itype = get_itype(itype)
+ if index is not None:
+ self._itype = get_itype(index)
+ if itype is not None:
+ self._itype = get_itype(itype)
if cast_policy not in _CAST_POLICIES:
raise ValueError(f"downcast_policy must be one of {_CAST_POLICIES}")
self._policy = cast_policy
- index = pd.Index([] if columns is None else columns)
- self._data = pd.Series(dtype='O', index=index.unique())
+ cols = pd.Index([] if columns is None else columns)
+ self._data = pd.Series(dtype='O', index=cols.unique())
if data is not None:
- self._init_insert_data(data)
+ self._init_insert_data(data, index=index)
# self._data contain nans at locations, where no
# data was present, but a column-name was given
@@ -109,14 +115,13 @@ class DictOfSeries:
# insert empty columns
if self._data.hasnans:
- e = pd.Series(dtype='O')
+ e = pd.Series(dtype='O', index=index)
for c in self.columns[self._data.isna()]:
self._insert(c, e.copy())
- self._data.name = '_data'
self._data.index.name = 'columns'
- def _init_insert_data(self, data):
+ def _init_insert_data(self, data, index=None):
""" Insert items of a iterable in self"""
data = list(data) if _is_iterator(data) else data
@@ -125,7 +130,8 @@ class DictOfSeries:
if isinstance(data, dict) or _is_dios_like(data):
for k in data:
if cols_empty or k in self.columns:
- self._insert(k, data[k])
+ s = pd.Series(data[k], index=index)
+ self._insert(k, s)
elif isinstance(data, pd.Series):
name = data.name or 0
@@ -133,7 +139,8 @@ class DictOfSeries:
# overwrite by columns
if data.name is None or data.name not in self.columns:
name = self.columns[0]
- self._insert(name, data)
+ s = pd.Series(data, index=index)
+ self._insert(name, s)
elif _is_list_like(data):
data = data if _is_nested_list_like(data) else [data]
@@ -145,7 +152,8 @@ class DictOfSeries:
raise ValueError(f"length of passed values is {len(data)}, columns imply {len(self.columns)}")
for i, c in enumerate(self.columns):
- self._insert(c, data[i])
+ s = pd.Series(data[i], index=index)
+ self._insert(c, s)
else:
raise ValueError(f"data must be some kind of iterable, type {type(data)} was given")
@@ -157,9 +165,13 @@ class DictOfSeries:
val = val.squeeze()
if not isinstance(val, pd.Series):
raise ValueError(f"Cannot insert frame-like with more than one column")
- elif not isinstance(val, pd.Series):
+
+ elif val is None:
val = pd.Series(val)
+ elif not isinstance(val, pd.Series):
+ raise TypeError(f"Only data of type pandas.Series can be inserted, passed was {type(val)}")
+
val = cast_to_itype(val, self.itype, policy=self._policy).copy(deep=True)
val.name = col
self._data.at[col] = val
--
GitLab