def main():
c = dltk_ai.DltkAiClient('xxx')
response = c.face_detection_image(
'../img/fd-actual-img.jpg') # it will return image data in bytes.
print(response)
save_img(response, "../img/fd-response.png")
response = c.face_detection_json(
'../img/fd-actual-img.jpg'
) # it will return the co-ordinates of detected faces
print(response)
response = c.eye_detection_image(
'../img/fd-actual-img.jpg') # it will return image data in bytes.
print(response)
save_img(response, "../img/fd-response.png")
response = c.eye_detection_json(
'../img/fd-actual-img.jpg'
) # it will return the co-ordinates of detected faces
print(response)
response = c.face_detection_image_core(
'../img/fd-actual-img.jpg') # it will return image data in bytes.
print(response)
save_img(response, "../img/fd-response.png")
response = c.face_detection_json_core(
'../img/fd-actual-img.jpg'
) # it will return the co-ordinates of detected licence plates.
print(response)
response = c.license_plate_detection_image(
'../img/lp-actual-img.jpg') # it will return image data in bytes.
print(response)
save_img(response, "../img/lp-response.png")
response = c.object_detection_image(
'../img/lp-actual-img.jpg') # it will return image data in bytes.
print(response)
save_img(response, "../img/lp-response.png")
response = c.object_detection_json(
'../img/lp-actual-img.jpg'
) # it will return the co-ordinates of detected licence plates.
print(response)
response = c.image_classification('../img/lp-actual-img.jpg')
print(response)
def main():
c = dltk_ai.DltkAiClient('xxx')
cluster_data_store_response = c.store('../csv/airoplane_data.csv', Dataset.TRAIN_DATA)
print(cluster_data_store_response)
cluster_data = cluster_data_store_response['fileUrl']
cluster_response = c.cluster("clustering", "KMeansClustering", cluster_data,
["Activity Period", "Operating Airline"], "spotflock", 2, "Clustering Model", True)
print(cluster_response)
cluster_job_status_response = c.job_status(cluster_response['data']['jobId'])
print(cluster_job_status_response)
cluster_job_output_response = c.job_output(cluster_response['data']['jobId'])
print(cluster_job_output_response)
pred_file = cluster_job_output_response['output']['clusterFileUrl']
response = c.download(pred_file)
print(response.text)
def main():
c = dltk_ai.DltkAiClient('xxx')
cluster_data_store_response = c.store('../csv/moon_data.csv',
Dataset.TRAIN_DATA)
print(cluster_data_store_response)
cluster_data = cluster_data_store_response['fileUrl']
cluster_response = c.cluster("clustering", "KMeansClustering",
cluster_data, ["X", "Y"], "scikit", 2,
"Clustering_Model", True, None)
print(cluster_response)
cluster_job_status_response = c.job_status(
cluster_response['data']['jobId'])
print(cluster_job_status_response)
cluster_job_output_response = c.job_output(
cluster_response['data']['jobId'])
print(cluster_job_output_response)
pred_file = cluster_job_output_response['output']['clusterFileUrl']
response = c.download(pred_file)
print(response.text)
#Loan Default Prediction
import dltk_ai
from dltk_ai.dataset_types import Dataset
#DLTK SDK requires Python 3.5 + . Go to https://dev.dltk.ai/ and create an app. On creation of an app, you will get an API Key.
c = dltk_ai.DltkAiClient('API Key')
#It stores Train and Test Files remotely. File upload API will return file storage locations from Cloud Storage in response.
train_file_store_response = c.store("path/to/train/file", Dataset.TRAIN_DATA)
train_data = train_file_store_response["fileUrl"]
#train_data file url
#'/spotflock-studio-prod/[email protected]/1551936734455-loan_train.csv'
#Upload Test File
test_file_store_response = c.store("path/to/test/file", Dataset.TEST_DATA)
test_data = test_file_store_response["fileUrl"]
#test_data file url
#'/spotflock-studio-prod/[email protected]/1551936725437-loan_test.csv'
#build model using NaiveBayesMultinomial algorithm.
train_response_NBM = c.train(
"classification", "NaiveBayesMultinomial", train_data, "Loan_Status",
["CurrentLoanAmount", "CreditScore", "MonthlyDebt", "Yr_Credit_His"],
"Loan Model - NaiveBayesMultinomial", "weka", 80, True)
#The train/predict jobs take some amount of time to be completed and so their status can be checked with this API.
train_job_status_response_NBM = c.job_status(
train_response_NBM["data"]["jobId"])
def main():
c = dltk_ai.DltkAiClient('xxx')
# train model
test_file_store_response = c.store('../csv/player_test.csv',
Dataset.TEST_DATA)
print(test_file_store_response)
test_data = test_file_store_response['fileUrl']
train_data_store_response = c.store('../csv/player_train.csv',
Dataset.TRAIN_DATA)
print(train_data_store_response)
train_data = train_data_store_response['fileUrl']
train_response = c.train("classification", "NaiveBayesMultinomial",
train_data, "player_activity",
["stamina", "challenges", "achievements"])
print(train_response)
train_job_status_response = c.job_status(train_response['data']['jobId'])
print(train_job_status_response)
train_job_output_response = c.job_output(train_response['data']['jobId'])
print(train_job_output_response)
model = train_job_output_response['output']['modelUrl']
predict_response = c.predict("classification", test_data, model, "weka")
print(predict_response)
predict_job_status_response = c.job_status(
predict_response['data']['jobId'])
print(predict_job_status_response)
predict_job_output_response = c.job_output(
predict_response['data']['jobId'])
print(predict_job_output_response)
pred_file = predict_job_output_response['output']['predFileUrl']
response = c.download(pred_file)
print(response.text)
# feedback model
# Feedback config should be same as train config except training percentage.
# Job id, model url, dataset url used in training a model is required to feedback any model.
job_id = train_response['data']['jobId']
# IMP: Ensure the dataset has all features and label used for training the model.
feedback_data_store_response = c.store('../csv/player_feedback.csv',
Dataset.TRAIN_DATA)
print(feedback_data_store_response)
feedback_data = feedback_data_store_response['fileUrl']
feedback_response = c.feedback("classification", "NaiveBayesMultinomial",
train_data, feedback_data, job_id, model,
"player_activity",
["stamina", "challenges", "achievements"])
print(feedback_response)
feedback_job_status_response = c.job_status(job_id)
print(feedback_job_status_response)
feedback_job_output_response = c.job_output(job_id)
print(feedback_job_output_response)
model = feedback_job_output_response['output']['modelUrl']
feedback_predict_response = c.predict("classification", test_data, model,
"weka")
print(feedback_predict_response)
feedback_predict_job_status_response = c.job_status(
predict_response['data']['jobId'])
print(feedback_predict_job_status_response)
feedback_predict_job_output_response = c.job_output(
predict_response['data']['jobId'])
print(feedback_predict_job_output_response)
pred_file = feedback_predict_job_output_response['output']['predFileUrl']
response = c.download(pred_file)
print(response.text)
def main():
c = dltk_ai.DltkAiClient('xxx') # put your app key here.
# REGRESSION Training
test_data = ""
train_data = ""
test_file_store_response = c.store('../csv/housing_test.csv',
Dataset.TEST_DATA)
print(test_file_store_response)
test_data = test_file_store_response['fileUrl']
train_data_store_response = c.store('../csv/housing_train.csv',
Dataset.TRAIN_DATA)
print(train_data_store_response)
train_data = train_data_store_response['fileUrl']
train_response = c.train("regression", "LinearRegression", train_data,
"SalePrice", ["LotShape", "Street"],
"Housing Price Model", "weka", 80,
True) # this is the configuration.
print(train_response)
train_job_status_response = c.job_status(train_response['data']['jobId'])
print(train_job_status_response)
train_job_output_response = c.job_output(train_response['data']['jobId'])
print(train_job_output_response)
model = train_job_output_response['output']['modelUrl']
predict_response = c.predict("regression", test_data, model, "weka")
print(predict_response)
predict_job_status_response = c.job_status(
predict_response['data']['jobId'])
print(predict_job_status_response)
predict_job_output_response = c.job_output(
predict_response['data']['jobId'])
print(predict_job_output_response)
pred_file = predict_job_output_response['output']['predFileUrl']
prediction_response = c.download(pred_file)
print(prediction_response.text)
# REGRESSION Feedback
job_id = train_response['data']['jobId']
# IMP: Ensure the dataset has all features and label used for training the model.
feedback_data_store_response = c.store('../csv/housing_feedback.csv',
Dataset.TRAIN_DATA)
print(feedback_data_store_response)
feedback_data = feedback_data_store_response['fileUrl']
feedback_response = c.feedback("regression", "LinearRegression",
train_data, feedback_data, job_id, model,
"SalePrice", ["LotShape", "Street"], "weka",
"Housing Price Model", 80, True)
print(feedback_response)
feedback_job_status_response = c.job_status(job_id)
print(feedback_job_status_response)
feedback_job_output_response = c.job_output(job_id)
print(feedback_job_output_response)
model = train_job_output_response['output']['modelUrl']
feedback_predict_response = c.predict("classification", test_data, model,
"weka")
print(feedback_predict_response)
feedback_predict_job_status_response = c.job_status(
predict_response['data']['jobId'])
print(feedback_predict_job_status_response)
feedback_predict_job_output_response = c.job_output(
predict_response['data']['jobId'])
print(feedback_predict_job_output_response)
pred_file = feedback_predict_job_output_response['output']['predFileUrl']
response = c.download(pred_file)
print(response.text)
def main():
c = dltk_ai.DltkAiClient('xxx')
response = c.sentiment_analysis('I am feeling good.')
print(response)
response = c.pos_tagger('I am Aniket.')
print(response)
import dltk_ai
client = dltk_ai.DltkAiClient(
'bb93840d-4578-4b2c-9bf4-9d056c44af65') # put your app key here.
# REGRESSION Training
test_file_store_response = client.store('../csv/rg_test.csv')
print(test_file_store_response)
test_data = test_file_store_response['fileUrl']
train_data_store_response = client.store('../csv/rg_train.csv')
print(train_data_store_response)
train_data = train_data_store_response['fileUrl']
train_response = client.train(
"classification", "LogisticRegression", train_data, 'Revenue.Grid',
['children', 'year_last_moved', 'Average.Credit.Card.Transaction'],
"Revenue_Grid_Model", "scikit", 80, True) # this is the configuration.
print(train_response)
train_job_status_response = client.job_status(train_response['data']['jobId'])
print(train_job_status_response)
train_job_output_response = client.job_output(train_response['data']['jobId'])
print(train_job_output_response)
model = train_job_output_response['output']['modelUrl']
predict_response = client.predict("classification",
test_data,
model,
"scikit",
features=[
'children', 'year_last_moved',
'Average.Credit.Card.Transaction'
])
print(predict_response)
predict_job_status_response = client.job_status(
