Series in Arkouda¶
Like Pandas, Arkouda supports Series. The purpose and intended functionality remains the same in Arkouda, but are configured to be based on arkouda.pdarrays.
- class arkouda.Series(data, name=None, index=None)[source]¶
One-dimensional arkouda array with axis labels.
- Parameters:
- Raises:
TypeError – Raised if index is not a pdarray or Strings object Raised if data is not a pdarray, Strings, or Categorical object
ValueError – Raised if the index size does not match data size
Notes
The Series class accepts either positional arguments or keyword arguments. If entering positional arguments,
- 2 arguments entered:
argument 1 - data argument 2 - index
- 1 argument entered:
argument 1 - data
If entering 1 positional argument, it is assumed that this is the data argument. If only ‘data’ argument is passed in, Index will automatically be generated. If entering keywords,
‘data’ (see Parameters) ‘index’ (optional) must match size of ‘data’
Features¶
Series support the majority of functionality offered by pandas.Series.
Lookup¶
- arkouda.Series.locate(self, key)¶
Lookup values by index label.
- Parameters:
key (int, pdarray, Index, Series, List, or Tuple) –
The key or keys to look up. This can be: - A scalar - A list of scalars - A list of lists (for MultiIndex) - A Series (in which case labels are preserved, and its values are used as keys)
Keys will be converted to Arkouda arrays as needed.
- Returns:
A Series containing the values corresponding to the key.
- Return type:
Lookup¶
- arkouda.Series.locate(self, key)¶
Lookup values by index label.
- Parameters:
key (int, pdarray, Index, Series, List, or Tuple) –
The key or keys to look up. This can be: - A scalar - A list of scalars - A list of lists (for MultiIndex) - A Series (in which case labels are preserved, and its values are used as keys)
Keys will be converted to Arkouda arrays as needed.
- Returns:
A Series containing the values corresponding to the key.
- Return type:
Sorting¶
- arkouda.Series.sort_index(self, ascending=True)¶
Sort the Series by its index.
- Parameters:
ascending (bool, default=True) – Whether to sort the index in ascending (default) or descending order.
- Returns:
A new Series sorted by index.
- Return type:
Head/Tail¶
- arkouda.Series.topn(self, n=10)¶
Return the top values of the Series.
- Parameters:
n (int, default=10) – Number of values to return. The default of 10 returns the top 10 values.
- Returns:
A new Series containing the top n values.
- Return type:
Value Counts¶
- arkouda.Series.value_counts(self, sort=True)¶
Return a Series containing counts of unique values.
- Parameters:
sort (bool, default=True) – Whether to sort the result by count in descending order. If False, the order of the results is not guaranteed.
- Returns:
A Series where the index contains the unique values and the values are their counts in the original Series.
- Return type:
Pandas Integration¶
- arkouda.Series.to_pandas(self)¶
Convert the series to a local PANDAS series
- Return type:
Series
- arkouda.Series.pdconcat(arrays, axis=0, labels=None)¶
Concatenate a list of Arkouda Series or grouped arrays, returning a local pandas object.
If a list of grouped Arkouda arrays is passed, they are converted to Series. Each grouping is a 2-tuple with the first item being the key(s) and the second the value.
If axis=1 (horizontal), each Series or grouping must have the same length and the same index. The index is converted to a column in the resulting DataFrame. If it is a MultiIndex, each level is converted to a separate column.
- Parameters:
arrays (List) – A list of Series or groupings (tuples of index and values) to concatenate.
axis (int, default=0) – The axis along which to concatenate: - 0 = vertical (stack into a Series) - 1 = horizontal (align by index into a DataFrame)
labels (Strings or None, optional) – Names to assign to the resulting columns in the DataFrame.
- Returns:
If axis=0: a local pandas Series
If axis=1: a local pandas DataFrame
- Return type: