Python in 1010data: Basic syntax

You can use Python to build query operations or ops in the 1010data Macro Language.

The <code> tag in Macro Language now supports the attribute language_="python", allowing your analysis to contain Python code. This means that table transformations, query operations, and metadata transformation can now be done with Python. The Python code is run in the accum. See <code> in the 1010data Reference Manual for more information about the <code> tag.

When you invoke <code language_="python">, the platform automatically imports numpy, pandas, sklearn, and ten. numpy is imported as np and pandas is imported as pd. ten is a 1010data-supplied module that is loaded into the Python session. ten is used to import data and metadata, get the number of rows, and set to temp tables or worksheets.

On entering the Python session from <code>, the session automatically gets the variables ops and table. ops represents the current state of query operations in effect up to the <code> tag. table is the base table, and will normally be default.lonely. On exiting <code>, the platform will examine the Python session for a variable named ops and extract it. The new ops variable will now represent the current state of the query. This allows the user to modify the query by modifying ops or replace the query entirely with a completely new set of ops.

For example, the end of your Python code can use the rebase method in ten and then set the value of ops as follows:

ops=ten.rebase(example_data_frame)

The DataFrame example_data_frame will then become a worksheet, and on exiting <code>, the DataFrame will effectively become the new state of the query.

When using the <code language_="python"> tag, it is good practice to put the Python functions and related operations inside of a CDATA tag.

<code language_="python">
<![CDATA[python code]]>
</code>

<base table="default.lonely"/>
<code language_="python">
<![CDATA[  
id = np.array([7685674,27363,7996,6943],dtype=np.int32)
col_names = ['Turquoise','Chartreuse','Lavender','Periwinkle']
ops = ten.rebase(pd.DataFrame({'ids':id,'color':col_names}))
]]>
</code>