def app_function():
    try:
        # Read account id and private from environment variables
        account_id = account['ID']
        private_key = ''
        with open('einstein_platform.pem', 'r') as f:
            for line in f:
                private_key += line

        # Set expiry time
        expiry = int(time.time()) + (15 * 60)

        # Generate an assertion using RSA private key
        assertion = jwt_helper.generate_assertion(account_id, private_key,
                                                  expiry)

        # Obtain oauth token
        token = token_generator.generate_token(assertion)
        response = token.json()

        # If there is no token print the response
        if 'access_token' not in response:
            raise ValueError(
                "Access token generation failed. Received reply: \"{}\"".
                format(response))
        else:
            # Collect the access token from response
            access_token = response['access_token']

        # Make a prediction call using image url
        prediction_url_response = prediction.predict_with_url(
            access_token, 'GeneralImageClassifier',
            'https://animalso.com/wp-content/uploads/2017/01/Siberian-Husky_7.jpg'
        )

        # Print prediction response
        pprint.pprint(prediction_url_response.json())

        # Make a prediction call using image file
        # prediction_file_response = prediction.predict_with_image_file(access_token, 'GeneralImageClassifier',
        #     '/path/to/image/file/Siberian-Husky.jpg')

        # Print prediction response
        # pprint.pprint(prediction_file_response.json())

    except Exception as e:
        traceback.print_exc()
Пример #2
0
def main():

    try:
        print("Hello")
        # Read account id and private from environment variables
        account_id = os.environ['EINSTEIN_VISION_ACCOUNT_ID']
        private_key = os.environ['EINSTEIN_VISION_PRIVATE_KEY'].decode(
            'string_escape')

        # Set expiry time
        expiry = int(time.time()) + (15 * 60)

        # Generate an assertion using RSA private key
        assertion = jwt_helper.generate_assertion(account_id, private_key,
                                                  expiry)
        # Obtain oauth token
        token = token_generator.generate_token(assertion)
        print("CC")
        response = token.json()
        print(response)
        # If there is no token print the response
        if 'access_token' not in response:
            raise ValueError(
                "Access token generation failed. Received reply: \"{}\"".
                format(response))
        else:
            # Collect the access token from response
            access_token = response['access_token']

        # Make a prediction call using image url
        prediction_url_response = prediction.predict_with_url(
            access_token, 'GeneralImageClassifier',
            'https://animalso.com/wp-content/uploads/2017/01/Siberian-Husky_7.jpg'
        )

        # Print prediction response
        pprint.pprint(prediction_url_response.json())

        # Make a prediction call using image file
        # prediction_file_response = prediction.predict_with_image_file(access_token, 'GeneralImageClassifier',
        #     '/path/to/image/file/Siberian-Husky.jpg')

        # Print prediction response
        # pprint.pprint(prediction_file_response.json())

    except Exception as e:
        traceback.print_exc()
Пример #3
0
def main():
    try:
        account_id = '*****@*****.**'  # Your Einstein Account ID
        private_key = open("einstein_platform.pem",
                           "r").read()  #Your Einstein Private Key file

        # Set expiry time
        expiry = int(time.time()) + (15 * 60)

        # Generate an assertion using RSA private key
        assertion = jwt_helper.generate_assertion(account_id, private_key,
                                                  expiry)

        # Obtain oauth token
        token = token_generator.generate_token(assertion)
        response = token.json()

        # If there is no token print the response
        if 'access_token' not in response:
            raise ValueError(
                "Access token generation failed. Received reply: \"{}\"".
                format(response))
        else:
            # Collect the access token from response
            access_token = response['access_token']

        #data_file='Rakesh_Sample.csv' #SFCC_Q1_2019_Recognition_Data.csv'
        data_file = 'TextFileForSentimentAnalysis.csv'
        data_file_with_sentiment = "Sentiment_of_" + data_file
        cleandata(data_file)
        df = pandas.read_csv(data_file)

        #Add three additional column to update sentiment value for the text
        df.insert(1, 'positive', '')
        df.insert(2, 'neutral', '')
        df.insert(3, 'negative', '')
        #df.insert(4,'word_count','')

        for index, row in df.iterrows():
            # Make a Sentiment prediction call

            print('Submited request for sentiment analysis of of text index #',
                  index)
            prediction_url_response = predictionSentiment.predict(
                access_token, row['citation'])
            print('Recieved sentiment for text index #', index)
            # Print prediction response
            resp = prediction_url_response.json()
            probabilities = resp['probabilities']

            for sentiment in probabilities:
                #pprint.pprint(sentiment)
                if sentiment['label'] == 'positive':
                    row['positive'] = sentiment['probability']
                elif sentiment['label'] == 'neutral':
                    row['neutral'] = sentiment['probability']
                elif sentiment['label'] == 'negative':
                    row['negative'] = sentiment['probability']

        df.to_csv(data_file_with_sentiment)
        textStatistics(data_file_with_sentiment)

    except Exception as e:
        traceback.print_exc()
Пример #4
0
def main():

    try:
        # Read account id and private from environment variables
        #account_id = os.environ['EINSTEIN_VISION_ACCOUNT_ID']
        #private_key = os.environ['EINSTEIN_VISION_PRIVATE_KEY'].decode('string_escape')

        account_id = '*****@*****.**'
        private_key = """
-----BEGIN RSA PRIVATE KEY-----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-----END RSA PRIVATE KEY-----
"""

        # Set expiry time
        expiry = int(time.time()) + (15 * 60)

        # Generate an assertion using RSA private key
        assertion = jwt_helper.generate_assertion(account_id, private_key,
                                                  expiry)

        # Obtain oauth token
        token = token_generator.generate_token(assertion)
        response = token.json()

        # If there is no token print the response
        if 'access_token' not in response:
            raise ValueError(
                "Access token generation failed. Received reply: \"{}\"".
                format(response))
        else:
            # Collect the access token from response
            access_token = response['access_token']

        # Upload the dataset to einstein.ai
        DS = dataset(access_token=access_token)
        #path = 'https://raw.githubusercontent.com/kaul-vineet/socialstudio-ml/master/data/intent_tagging.csv'
        #response = DS.create_intent_dataset(path)
        #print(json.dumps(response, indent=4, sort_keys=True))

        # Train the model on einstein.ai []
        id = '1127772'
        #DS = dataset(access_token=access_token)
        response = DS.train_dataset(id)
        #if('available' in response):
        #    print(json.dumps(response, indent=4, sort_keys=True))
        #else:
        #    print('Response status ok?: ' + str(response.ok))
        #    print(json.dumps(response.text, indent=4, sort_keys=True))

        # Check the model training status on einstein.ai []
        id = 'YRVFEBIDWGX4I6EBKDOFU5KRQM'
        #response = DS.get_train_status(id)
        #print(json.dumps(response, indent=4, sort_keys=True))
        #data = json.loads(json.dumps(response))
        #print ('************ THE MODEL TRAINING IS IN PROGRESS ************')
        #while data['status'] != 'SUCCEEDED':
        #    print ('THE MODEL STATUS IS :' + data['status'])
        #    time.sleep(30)
        #else:
        #    print ('THE MODEL STATUS IS :' + data['status'] + ' WITH LEARNING RATE OF ' + str(data['learningRate']))

        # Check the predictions on einstein.ai []
        model_id = 'YRVFEBIDWGX4I6EBKDOFU5KRQM'
        document = 'hey guys, im a black trans creative named wondy!! i make art, unfortunately my account was suspended and i lost my 3.5k following and clientele :( please retweet this post so i can get my product back out there as this is my income!! any support is phenomenal'
        predict = prediction(access_token=access_token)
        response = predict.predict_social_tag(document, model_id)
        probabilities = response['probabilities']
        max_prob = 0
        max_tag = ''
        for x in probabilities:
            if max_prob < x['probability'] * 100:
                max_prob = x['probability'] * 100
                max_tag = str(x['label'])
        print('There is ' + str(max_prob) + ' probability that this is ' +
              max_tag + ' post.')

    except Exception as e:
        traceback.print_exc()