I am trying to find the list of unique colours from the CSV downloaded from https://data.cityofnewyork.us/Environment/2018-Central-Park-Squirrel-Census-Squirrel-Data/vfnx-vebw
The following does not work:
data = pandas.read_csv("2018_Central_Park_Squirrel_Census_-_Squirrel_Data.csv")
fur_color_col = data["Primary Fur Color"]
print(data[not pandas.isna(data["Primary Fur Color"])]["Primary Fur Color"].unique())
The error is:
Traceback (most recent call last):
File "/Users/arvind.avinash/PycharmProjects/AdHoc/main.py", line 6, in <module>
print(data[not pandas.isna(data["Primary Fur Color"])]["Primary Fur Color"].unique())
File "/Users/arvind.avinash/PycharmProjects/AdHoc/venv/lib/python3.9/site-packages/pandas/core/generic.py", line 1537, in __nonzero__
raise ValueError(
ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().
The following works:
data = pandas.read_csv("2018_Central_Park_Squirrel_Census_-_Squirrel_Data.csv")
fur_color_col = data["Primary Fur Color"]
print(data[pandas.isna(data["Primary Fur Color"]) != True]["Primary Fur Color"].unique())
and outputs:
['Gray' 'Cinnamon' 'Black']
Why does not bool
not work while bool != True
works in this case?
CodePudding user response:
Because non
for arrays (in pandas or numpy) is operator ~
, so need:
print(data[~pandas.isna(data["Primary Fur Color"])]["Primary Fur Color"].unique())
For compare in arrays (in pandas or numpy) are used same operators like in pure python, so != True
working well.
Or is possible use Series.notna
:
print(data.loc[data["Primary Fur Color"].notna(), "Primary Fur Color"].unique())
CodePudding user response:
The reason that not
does not work, is that it is defined by the language spec to return a single boolean value, which does not make much sense with a Pandas series.
The operator not yields True if its argument is false, False otherwise.
The !=
operator has no such limitations, which means that Pandas is free to define it as an element-by-element comparison.