cmd-commands.txt

MiWO Commands

Anaconda Prompt -> Run as Administrator

1.cd ../..
2. (base) C:\>cd Users/Sonalika/Documents
3. (base) C:\Users\Sonalika\Documents>.\rasabot\Scripts\activate
4. (rasabot) (base) C:\Users\Sonalika\Documents>cd A-RASA/miwo
5. (rasabot) (base) C:\Users\Sonalika\Documents\A-RASA\MiWo>rasa run -m models --endpoints endpoints.yml --port 5002 --credentials credentials.yml

(rasa x) 

(not required) OPEN ANOTHER TERMINAL - 
1. to 4. same as above
5. (rasabot) (base) C:\Users\Sonalika\Documents\A-RASA\MiWo>python -m rasa_sdk --actions actions


OPEN ANOTHER TERMINAL - 
1. to 4. same as above
5. (rasabot) (base) C:\Users\Sonalika\Documents\A-RASA\MiWo>python Voice_bot.py run


OPEN ANOTHER TERMINAL - 
1. to 4. same as above
5. (rasabot) (base) C:\Users\Sonalika\Documents\A-RASA\MiWo>docker run -p 8000:8000 rasa/duckling



(not required) 
How to extract the information from the excel sheet and display in the output

Add the given data to nlu.md to accept what user says,

## intent:ask_info
- list out the [email]{"entity":"column","value":"email"} of the user who's occupation is [trainer]{"entity":"occup","value":"trainer"}
- list out the [name]{"entity":"column","value":"name"} of the user who's occupation is [doctor]{"entity":"occup","value":"doctor"}
- tell me the [mobile number]{"entity":"column","value":"number"} of the user whose occupation is [software developer]{"entity":"occup","value":"software developer"}
this is intent with the entities that will be extracted from the user input based on which information will be extracted from the user’s existing excel sheet. Add the intent name and entities in the domain file to register them.

intents:
- ask_info
entities:
- column
- occup
Based on these entities extracted, create an action that will be called when the user will ask for the information to extract.
Add the given line of code in to the actions.py script to extract the data,

class ActionFetchData(Action):
def name(self) -> Text:
    return "action_fetch_data"

def run(self, dispatcher: CollectingDispatcher,
        tracker: Tracker,
        domain: Dict[Text, Any]) -> List[Dict[Text, Any]]:
    output=FetchData(tracker.latest_message['entities'][0]['value'],
                     tracker.latest_message['entities'][1]['value'])
    dispatcher.utter_message(text="This is the data that you asked for, \n{}".format(",".join(output)))

    return []
In the above run method, you will see a function is called which will be created in the other script to read the data from the excel sheet,

def FetchData(column,occupation):
    if os.path.isfile("user_data.xlsx"):
        df=pd.read_excel("user_data.xlsx")
        data=df[column][df["occupation"]==occupation]
        return data.to_list()
    else:
        return ["There is no data. "]
Once, you have created the function for reading the data as per the user input. Also, add the stories in the stories.md file so as to follow the path,

## fetch info
* ask_info{"column":"email","occup":"trainer"}
  - action_fetch_data

 
Also add this action name to domain file to register it,

actions:
- action_fetch_data

pip install SpeechRecognition PyAudio
pip install gtts
conda install mpg321