machine learning - TensorFlow/Inception Precision and Recall - Correct Nomenclature? -

the script provided in tensorflow models repo evaluating inception calculates calls precision , recall follows:

top_1_op = tf.nn.in_top_k(logits, labels, 1) top_5_op = tf.nn.in_top_k(logits, labels, 5) 

...

top_1, top_5 = sess.run([top_1_op, top_5_op]) count_top_1 += np.sum(top_1) count_top_5 += np.sum(top_5) 

...

precision_at_1 = count_top_1 / total_sample_count recall_at_5 = count_top_5 / total_sample_count 

now, understand it, precision , recall specific , distinct calculations. not code above doing same thing labels precision , recall, me, looks it's calculating accuracy (true positives + true negatives) / total population.

am misunderstanding definition of these terms in context?