def model(id):
model = mlModels.query.get(id)
form = PredictForm()
if form.validate_on_submit():
task = predictions(data=form.data.data.filename, model_id=id)
db.session.add(task)
db.session.commit()
s3Client.put_object(Bucket=S3Name,
Key=str(task.creationdate) +
form.data.data.filename,
Body=form.data.data)
SQS.send_message(QueueUrl=predictQueue,
MessageBody=json.dumps({'id': task.pred_id}))
return redirect(url_for('results', id=task.pred_id))
data = json.loads(model.data)
dictionary = json.loads(model.dict)
return render_template(
'model.html',
form=form,
model=model,
date=model.creationdate.strftime("%m/%d/%Y, %H:%M:%S"),
layers=dictionary['layers'],
sampling=dictionary['sampling'],
custom=dictionary['custom'],
customDesc=dictionary['model'],
dataType=data['dataType'])
def demo_heart(request):
context = {
"model": None,
"predictform": None,
"predictionresult": None,
"input": None,
}
model = get_object_or_404(ClassificationModel, pk=2)
context["model"] = model
predictform = PredictForm(model.variables)
context["predictform"] = predictform
if request.method == "POST":
predictform = PredictForm(model.variables, request.POST)
if predictform.is_valid():
cd = predictform.cleaned_data
prediction_result = prediction(cd, model)
if prediction_result == False:
messages.warning(request,
"Something went wrong with the prediction.")
else:
context["predictionresult"] = prediction_result
context["predictform"] = predictform
context["input"] = cd
else:
messages.warning(request,
"Something went wrong with the prediction.")
return render(request, "app/predict.html", context)
def text_predict(request):
if request.method == 'POST':
form = PredictForm(request.POST)
if form.is_valid():
my_dict = request.POST.dict()
result = predict(my_dict)
print(result)
messages.success(request, {result})
form = PredictForm()
return render(request, 'predict_form.html', {'form': form})
def predict():
result_LinearSVC = ""
result_NuSVC = ""
result_DecisionTree = ""
result_sentiment = ""
f1_LinearSVC, acc_LinearSVC = None, None
f1_NuSVC, acc_NuSVC = None, None
f1_DecisionTree, acc_DecisionTree = None, None
form = PredictForm()
if form.validate_on_submit():
result_LinearSVC = prediction(form.parent_text.data,
form.child_text.data, 'LinearSVC')
f1_LinearSVC, acc_LinearSVC = cross_validation('LinearSVC')
result_NuSVC = prediction(form.parent_text.data, form.child_text.data,
'NuSVC')
f1_NuSVC, acc_NuSVC = cross_validation('NuSVC')
result_DecisionTree = prediction(form.parent_text.data,
form.child_text.data, 'DecisionTree')
f1_DecisionTree, acc_DecisionTree = cross_validation('DecisionTree')
result_sentiment = sentiment_predict(form.parent_text.data,
form.child_text.data)
return render_template('predict.html',
result_LinearSVC=result_LinearSVC,
result_NuSVC=result_NuSVC,
result_DecisionTree=result_DecisionTree,
form=form,
f1_LinearSVC=f1_LinearSVC,
f1_DecisionTree=f1_DecisionTree,
f1_NuSVC=f1_NuSVC,
acc_LinearSVC=acc_LinearSVC,
acc_DecisionTree=acc_DecisionTree,
acc_NuSVC=acc_NuSVC,
result_sentiment=result_sentiment)
return render_template('predict.html',
result_LinearSVC=result_LinearSVC,
result_NuSVC=result_NuSVC,
result_DecisionTree=result_DecisionTree,
form=form,
f1_LinearSVC=f1_LinearSVC,
f1_DecisionTree=f1_DecisionTree,
f1_NuSVC=f1_NuSVC,
acc_LinearSVC=acc_LinearSVC,
acc_DecisionTree=acc_DecisionTree,
acc_NuSVC=acc_NuSVC,
result_sentiment=result_sentiment)
def predict():
form = PredictForm()
if request.method == 'POST':
if form.validate_on_submit():
key_array = []
prediction_array = []
for key in request.files.keys():
key_array.append(key)
images = request.files.getlist(key_array[0])
for img in images:
filename = os.path.join(app.config['UPLOAD_FOLDER'],
save_file(img))
filename_array = filename.split('\\')
name = filename_array[len(filename_array) -
1].split('.')[0].split('_')[0]
prediction_object = {
'name': name,
'prediction': predict_file(filename, form.project.data)
}
prediction_array.append(prediction_object)
return render_template('predictions.html',
prediction=prediction_array)
return render_template('predict.html', title='Predict', form=form)
def predictdata():
form = PredictForm.new()
if form.validate_on_submit():
assets_dir = os.path.join(os.path.dirname(app.instance_path), 'assets')
input_data = form.datasheet.data
input_filename = secure_filename(input_data.filename)
model_name = form.model_name.data
input_data.save(os.path.join(assets_dir, 'tempdata', input_filename))
results_array = easy_ai_prediction.prediction(input_filename,
model_name)
session['results'] = results_array.tolist()
os.remove('C:/Users/Matt/Documents/graymatter-flask/assets/tempdata/' +
input_filename)
return redirect(url_for('results'))
return render_template('predictdata.html', title='Predict Data', form=form)
def PredictHeartDisease():
form = PredictForm()
return render_template('predict.html', title='Predict', form=form)