From 0432299cf59b8de00cedc5f8b44c5d4de62ea042 Mon Sep 17 00:00:00 2001
From: Bert Palm <bert.palm@ufz.de>
Date: Wed, 15 Apr 2020 02:44:32 +0200
Subject: [PATCH] simplified init a bit
---
dios/dios.py | 65 +++++++++++++++++++++++++++-------------------------
1 file changed, 34 insertions(+), 31 deletions(-)
diff --git a/dios/dios.py b/dios/dios.py
index ad16abd..032113b 100644
--- a/dios/dios.py
+++ b/dios/dios.py
@@ -100,20 +100,23 @@ class DictOfSeries:
self._policy = cast_policy
cols = pd.Index([] if columns is None else columns)
- self._data = pd.Series(dtype='O', index=cols.unique())
+ if not cols.is_unique:
+ raise ValueError("columns must be unique")
+ self._data = pd.Series(dtype='O', index=cols)
if data is not None:
- self._init_insert_data(data, index=index)
+ self._init_insert_data(data, columns, index)
- # self._data contain nans at locations, where no
- # data was present, but a column-name was given
+ # Note: self._data still contain nans at all positions
+ # where no data was present, but a column-name was given
if itype is None:
self._itype = _find_least_common_itype(self._data.dropna())
if not self._itype.unique:
_throw_MixedItype_err_or_warn(self.itype)
- # insert empty columns
+ # finalise data: insert empty
+ # columns at nan positions
if self._data.hasnans:
e = pd.Series(dtype='O', index=index)
for c in self.columns[self._data.isna()]:
@@ -121,41 +124,41 @@ class DictOfSeries:
self._data.index.name = 'columns'
- def _init_insert_data(self, data, index=None):
+ def _init_insert_data(self, data, columns, index):
""" Insert items of a iterable in self"""
- data = list(data) if _is_iterator(data) else data
- cols_empty = self.columns.empty
+ if _is_iterator(data):
+ data = list(data)
- if isinstance(data, dict) or _is_dios_like(data):
- for k in data:
- if cols_empty or k in self.columns:
- s = pd.Series(data[k], index=index)
- self._insert(k, s)
+ if _is_dios_like(data) or isinstance(data, dict):
+ if columns is None:
+ pass # data is dict-like
+ else:
+ data = {k: data[k] for k in data if k in columns}
elif isinstance(data, pd.Series):
name = data.name or 0
- if not cols_empty:
- # overwrite by columns
- if data.name is None or data.name not in self.columns:
- name = self.columns[0]
- s = pd.Series(data, index=index)
- self._insert(name, s)
+ if columns is not None and len(columns) > 0:
+ name = self.columns[0]
+ data = {name: data}
+
+ elif _is_nested_list_like(data):
+ if columns is None:
+ data = {i: d for i, d in enumerate(data)}
+ elif len(data) == len(columns):
+ data = dict(zip(self.columns, data))
+ else:
+ raise ValueError(f"{len(columns)} columns passed, data implies {len(data)} columns")
elif _is_list_like(data):
- data = data if _is_nested_list_like(data) else [data]
-
- if cols_empty:
- self._data = pd.Series(dtype='O', index=pd.RangeIndex(len(data)))
+ name = 0 if columns is None or len(columns) < 1 else self.columns[0]
+ data = {name: data}
- elif len(data) != len(self.columns):
- raise ValueError(f"length of passed values is {len(data)}, columns imply {len(self.columns)}")
-
- for i, c in enumerate(self.columns):
- 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")
+ raise TypeError("data type not understood")
+
+ for k in data:
+ self._insert(k, pd.Series(data[k], index=index))
def _insert(self, col, val):
"""Insert a fresh new value as pd.Series into self"""
@@ -167,7 +170,7 @@ class DictOfSeries:
raise ValueError(f"Cannot insert frame-like with more than one column")
elif val is None:
- val = pd.Series(val)
+ val = pd.Series()
elif not isinstance(val, pd.Series):
raise TypeError(f"Only data of type pandas.Series can be inserted, passed was {type(val)}")
--
GitLab