Pulling signal's variable name

[Mohammad Reza Babaei]

Hi,

Can users access the signal's variable name in the data lab? It would be great to have it as a column when calling spy.search(worksheet_url ...).

For example:

[Mark Derbecker]

Hi Mohammad,

When you include the "all_properties=True" argument in spy.search(), you will receive a "Formula Parameters" column that has the variables as a list of strings in the form "variable_name=item_id" like so:

To get exactly the DataFrame you're asking for in your original post, you can then use the following DataFrame manipulation code and another call to spy.search():

variables_df = pd.DataFrame([{'Variable Name': f"${fp.split('=')[0]}", 'ID': fp.split('=')[1]}
                             for fp in search_df.iloc[0]['Formula Parameters']])
merged_df = spy.search(variables_df)
merged_df = merged_df[['ID', 'Name', 'Type']]
merged_df = merged_df.merge(variables_df, on='ID')
merged_df

It'll look like this:

[Mohammad Reza Babaei]

That's great! Thanks, Mark.

I wanted to start writing a formula using Stored Signals in Data Lab. In this case, I will get the following:

I tried to write a function for converting "Name" to "Variable Name":

from re import split

def create_formula_variable_name(names):
    formula_name = []
    for name_i in names:
        name_i_lower = name_i.lower()
        name_i_list = split("_| | ", name_i_lower)
        f_name = '$'
        
        cnt = 0
        if len(name_i_list) > 1:
            for i in name_i_list:
                f_name += i[0]
        else:
            if len(name_i_list[0]) > 4:
                f_name += name_i_list[0][0]
            else:
                f_name += name_i_list[0]

        formula_name.append(f_name)

    return formula_name

However, this is not a perfect solution.

[Mark Derbecker]

It is better to correlate the calculation variables with the stored signals by using the ID -- the name of the variable can be anything, so it is not a reliable link between the two.

[Mohammad Reza Babaei]

Thank you very much Mark! I will do that.