Using tabular inputs to build a model

When the number of users or appliances is high, it can be difficult to create a model using Python scripts. Therefore, RAMP allows you to create inputs in tabular format (.xlsx). On the other hand, it is still possible to use Python to generate a large tabular file with default parameter values, which can then be more easily customised. In this example, we show a possible utilisation of this functionality.

from ramp import User, Appliance, UseCase, get_day_type
import pandas as pd

As a first step, one must create User classes and assign Appliances to the user class without assigning detailed appliance characteristics. Hence, users and their appliances are added to a UseCase.

Building a tabular file populated with default data

# Defining a dict of users with their appliances

user_app = {"household": ["light", "tv"], "school": ["light", "computer"]}

# creating a UseCase class to create the database
use_case = UseCase()

# assinging the appliances to users
for user, apps in user_app.items():
    user_instance = User(user_name=user)

    for app in apps:
        app_instance = user_instance.add_appliance(name=app)
        app_instance.windows()

    use_case.add_user(user_instance)

/home/fl/GitHub-repos/RAMP/ramp/core/core.py:1198: UserWarning: No windows is declared, default window of 24 hours is selected
  warnings.warn(

Once the Users and Appliances are added to the use_case instance, the model user can get a pd.DataFrame or an .xlsx file of all the data with the default values.

Exporting the database

# getting the dataframe
use_case.export_to_dataframe()

	user_name	num_users	user_preference	name	number	power	num_windows	func_time	time_fraction_random_variability	func_cycle	...	cw32_start	cw32_end	r_c3	window_1_start	window_1_end	window_2_start	window_2_end	window_3_start	window_3_end	random_var_w
0	household	1	0	light	1	0.0	1	0	0	1	...	0	0	0	0	1440	0	0	0	0	0
1	household	1	0	tv	1	0.0	1	0	0	1	...	0	0	0	0	1440	0	0	0	0	0
2	school	1	0	light	1	0.0	1	0	0	1	...	0	0	0	0	1440	0	0	0	0	0
3	school	1	0	computer	1	0.0	1	0	0	1	...	0	0	0	0	1440	0	0	0	0	0

4 rows × 51 columns

# Printing out the database to an .xlsx file
use_case.save("example_excel_usecase")

Once the function is used, an .xlsx file will be created in the given path. Now, you can easily fill out the information in the .xlsx file and load the data into the model database as detailed below.

Loading the database

# loading data

use_case = UseCase()  # creating a new UseCase instance
use_case.load("example_excel_usecase_filled.xlsx")

Generating load profiles

Once the database is loaded, the user can continue with the normal analysis, for instance, generating aggregated profiles

n_days = 30
date_start = "2020-01-01"
use_case.date_start = date_start
use_case.initialize(num_days=n_days, force=True)
use_case.generate_daily_load_profiles()

You will simulate 30 day(s) from 2020-01-01 00:00:00 until 2020-01-31 00:00:00

array([0.   , 0.   , 0.   , ..., 0.002, 0.002, 0.002])

profiles = pd.DataFrame(
    data=use_case.generate_daily_load_profiles(flat=True),
    index=pd.date_range(start=date_start, periods=1440 * n_days, freq="T"),
)

profiles.plot(title="Usecase")

<Axes: title={'center': 'Usecase'}>

Generating load profiles for the single users of the usecase

for user in use_case.users:
    user_profiles = []
    for day_idx, day in enumerate(use_case.days):
        profile = user.generate_aggregated_load_profile(
            prof_i=day_idx,
            peak_time_range=use_case.peak_time_range,
            day_type=get_day_type(day),
        )

        user_profiles.extend(profile)

    profiles = pd.DataFrame(
        data=user_profiles,
        index=pd.date_range(start=date_start, periods=1440 * n_days, freq="T"),
    )

    profiles.plot(title=user.user_name)

Link to the jupyter notebook file.