Xtykutl

I have a function for running a logistic regression model. I would like to permanently save the dataframe generated within (cf) and amend it's name.

def model(ind, dep):

    global cf

    ind.fillna(0)

    #some modelling code



    #create confuson matix



    cf = pd.DataFrame(confusion_matrix(y_train, y_pred))

    cf.index = models

    cf.columns = models

    print(cf)

    cf.plot.barh()





    "cf_" + str(ind) = cf



    return "cf_" + str(ind)



model(X_tv, y_combo)

model(X_tv_chan, y_combo)

I get this error

File "", line 64
"cf_" + str(ind) = coeff
^
SyntaxError: can't assign to operator

is there something wrong with how I am trying to create the dataframe?

"cf_" + str(ind) = cf

Yes. This line doesn't create a new variable:

"cf_" + str(ind) = cf

Strings are immutable. You can't "assign a dataframe to a string" either, I'm not even sure what this is attempting to achieve. It's also good practice to avoid global variables.

Just return your dataframe and assign to an explicit variable name. If you are set on changing your variable name, use a dictionary and dict.pop:

def model(ind, dep):

    # ...

    cf = pd.DataFrame(confusion_matrix(y_train, y_pred))

    cf.index = models

    cf.columns = models

    # ...

    return cf



dfs = {}



dfs['ind'] = model(X_tv, y_combo)  # identify dataframe with key 'ind'

dfs['cf_ind'] = dfs.pop('ind')     # rename identifier to 'cf_ind'