View source for Python Pandas Library

[[category:Python integration]]
''new to [[Analytica 7.0]]''

The Python Pandas library contains functions to assist with using the Pandas library (a common Python library) from within your Analytica model.

== PyExplodeDataFrame(dataFrame'',  columns, rows, separator, byPosition'') ==

Use this function to convert a Pandas DataFrame into an Analytica multidimensional table, optionally matching each DataFrame axis to provided Analytica Indexes either by label (default) or by position.

;«dataFrame»
:The Pandas DataFrame to convert into an Analytica array.

;«columns»
:One or more Analytica Indexes corresponding to the «dataFrame»'s column axis, ordered from outermost to innermost level.
:When specified, the number of provided Indexes must match the number of MultiIndex levels, otherwise an error is thrown.
:If the column axis is a MultiIndex and «columns» is provided, each MultiIndex level becomes a separate Analytica dimension.
:If «columns» is omitted, a local index is created from the labels of the «dataFrame»’s column axis. For a MultiIndex, the levels are flattened into a single set of labels joined by «separator».

;«rows»
:One or more Analytica Indexes corresponding to the «dataFrame»'s row axis, ordered from outermost to innermost level.
:When specified, the number of provided Indexes must match the number of MultiIndex levels, otherwise an error is thrown.
:If the row axis is a MultiIndex and «rows» is provided, each MultiIndex level becomes a separate Analytica dimension.
:If «rows» is omitted, a local index is created from the labels of the «dataFrame»’s row axis. For a MultiIndex, the levels are flattened into a single set of labels joined by «separator».

;«separator»
:Used only when a local index is created for a MultiIndex axis; joins the labels from multiple levels into a single flat label. Ignored when the axis is not a MultiIndex or when explicit Indexes for an axis are provided.

;«byPosition»
*False — Matches «dataFrame» axes to provided Analytica Indexes by label.
*:The order of Analytica Index items determines the order of values in the result — data is reordered if labels are out of order.
*:Extra Analytica Index items not present in the «dataFrame» labels are null padded; labels in the «dataFrame» but not in the Analytica Index are dropped.
*:Matching is applied independently to each axis and to each MultiIndex level.
*:NOTE: Each Index's identifier must exactly match a level name in the MultiIndex
*:Example: <code>DataFrame ['One','tWo','thRee'] → [1, 2, 3]</code> with Analytica Index <code>['thRee','tWo','one']</code> gives <code>[3, 2, Null]</code>.
*:The Analytica Index label 'one' is not found and is therefore nullpadded. The label 'One' in «dataFrame» is not found in the Analytica Index and is therefore dropped.
*True — Matches «dataFrame» axes to Analytica Indexes by position, ignoring label names.
*:The first provided Analytica Index corresponds to the first MultiIndex level, the second to the second level, etc.
*:Each Index length must exactly match the length of the corresponding MultiIndex level, or an error is thrown.

== See also ==
* [[Using Python from Analytica]]
* [[PyExplode]]
* [[PyArray]]