create_view() | Teradata Package for Python - create_view() - Teradata Package for Python

Use the create_view() function to create a view from the DataFrame object in the specified schema. As Teradata Package for Python creates views, internally for operations, which will be garbage collected during remove_context(), using this helps persist the DataFrame as a view.

Example setup

Load the data to run the example.

>>> load_example_data("antiselect", ["antiselect_input"])
>>> antiselect_input = DataFrame.from_table("antiselect_input")
>>> antiselect_input

Output

        orderid orderdate       priority  quantity       sales  discount        shipmode            custname province   region     custsegment          prodcat
rowids
49          293  12/10/01           high        49  10123.0200      0.07  delivery truck        barry french  nunavut  nunavut        consumer  office supplies
97          613  11/06/17           high        12     93.5400      0.03     regular air        carl jackson  nunavut  nunavut       corporate  office supplies
85          515  10/08/28  not specified        19    394.2700      0.08     regular air      carlos soltero  nunavut  nunavut        consumer  office supplies
86          515  10/08/28  not specified        21    146.6900      0.05     regular air      carlos soltero  nunavut  nunavut        consumer        furniture
1             3  10/10/13            low         6    261.5400      0.04     regular air  muhammed macintyre  nunavut  nunavut  small business  office supplies
50          293  12/10/01           high        27    244.5700      0.01     regular air        barry french  nunavut  nunavut        consumer  office supplies
80          483  11/07/10           high        30   4965.7595      0.08     regular air       clay rozendal  nunavut  nunavut       corporate       technology

Filter the data based on quantity.

>>> anti_df = antiselect_input[antiselect_input.quantity < 30]
>>> anti_df

Output

        orderid orderdate       priority  quantity   sales  discount     shipmode            custname province   region     custsegment          prodcat
rowids
97          613  11/06/17           high        12   93.54      0.03  regular air        carl jackson  nunavut  nunavut       corporate  office supplies
86          515  10/08/28  not specified        21  146.69      0.05  regular air      carlos soltero  nunavut  nunavut        consumer        furniture
85          515  10/08/28  not specified        19  394.27      0.08  regular air      carlos soltero  nunavut  nunavut        consumer  office supplies
1             3  10/10/13            low         6  261.54      0.04  regular air  muhammed macintyre  nunavut  nunavut  small business  office supplies
50          293  12/10/01           high        27  244.57      0.01  regular air        barry french  nunavut  nunavut        consumer  office supplies

Run Antiselect on filtered data. This will create temporary view which will be garbage collected.

>>> obj = Antiselect(data=anti_df, exclude=['rowids', 'orderdate', 'discount', 'province', 'custsegment'])

Get the view name that is internally created by teradataml to store the result of Antiselect.

>>> obj.result.db_object_name

Output

'"<schema_name>"."ml__td_sqlmr_out__1752582812690000"'

Check the output of Antiselect.

>>> obj.result

Output

orderid       priority  quantity   sales     shipmode            custname   region          prodcat
0      613           high        12   93.54  regular air        carl jackson  nunavut  office supplies
1      515  not specified        21  146.69  regular air      carlos soltero  nunavut        furniture
2      515  not specified        19  394.27  regular air      carlos soltero  nunavut  office supplies
3      293           high        27  244.57  regular air        barry french  nunavut  office supplies
4        3            low         6  261.54  regular air  muhammed macintyre  nunavut  office supplies

Describe the resultant DataFrame.

>>> df = obj.result.describe() # This will create a temporary view.

Get the view name.

>>> df.db_object_name

Output

'"<schema_name>"."ml__td_sqlmr_out__1752585435339977"'

Check the output of describe.

>>> df

Output

    ATTRIBUTE            StatName   StatValue
0   orderid             MAXIMUM  613.000000
1   orderid  STANDARD DEVIATION  245.016734
2   orderid     PERCENTILES(25)  293.000000
3   orderid     PERCENTILES(50)  515.000000
4  quantity               COUNT    5.000000
5  quantity             MINIMUM    6.000000
6  quantity             MAXIMUM   27.000000
7  quantity                MEAN   17.000000
8  quantity  STANDARD DEVIATION    8.154753
9  quantity     PERCENTILES(25)   12.000000

Example: Persist the view which can be accessed across sessions

>>> df_new = df.create_view(view_name="antiselect_describe_view")
>>> df_new

Output

    ATTRIBUTE            StatName   StatValue
0  quantity             MAXIMUM   27.000000
1  quantity  STANDARD DEVIATION    8.154753
2  quantity     PERCENTILES(25)   12.000000
3  quantity     PERCENTILES(50)   19.000000
4     sales               COUNT    5.000000
5     sales             MINIMUM   93.540000
6   orderid               COUNT    5.000000
7   orderid             MINIMUM    3.000000
8   orderid             MAXIMUM  613.000000
9   orderid                MEAN  387.800000

Get the view name.

>>> df_new.db_object_name # "<schema_name>" is user connected database.

Output

'"<schema_name>"."antiselect_describe_view"'