The Taskframe Class: Exporting Labels

Once your team has finished annotating your Tasks, you have a few options to export the results.

Quick Summary

# Assuming a finished Taskframe:
tf = Taskframe.retrieve("xxxx")

# export to a CSV:
tf.to_csv("local/path/results.csv")

# export to a Pandas Dataframe:
df = tf.to_dataframe()

# Include the resulting labels on a new column of an existing dataframe:
df_with_label_column = tf.merge_to_dataframe(df)

to_csv

Export your results as a CSV Signature:

def to_csv(self, path):

Parameters:

• path: local path where the CSV will be saved.

Returns: None

Example:

# Assuming a finished Taskframe:
tf = Taskframe.retrieve("xxxx")
tf.to_csv("local/path/results.csv")

to_dataframe

Generate a Pandas dataframe with the results. Signature:

def to_dataframe(self):

Parameters: None

Returns: A pandas DataFrame with the results.

Example:

# Assuming a finished Taskframe:
tf = Taskframe.retrieve("xxxx")
df = tf.to_dataframe()

merge_to_dataframe

If your input Dataset is a Dataframe, you can merge your result labels as a new column in your original Dataframe. Requires that you had submitted custom_id to be able to join rows. Signature:

merge_to_dataframe(self, dataframe, custom_id_column):

Parameters:

• dataframe: the dataframe to which we will add a new labels column
• custom_id_colum: the name of the column containing unique identifiers (should be the same as provided when dataset was initially submitted)

Returns: a copy of the input dataframe with an extra label column containing the results.

Example:

df = pd.DataFrame(...)

tf.add_dataset_from_dataframe(df)
tf.submit()

# Your workers annotate, once it's done:
df = pd.DataFrame(...)
df_with_labels = tf.merge_to_dataframe(df)