simio-lisa 1.1.0

Beautiful Simio: Importing data and powering viz.

simio-lisa

Python package of processing tools for Simio models saved as .simproj

How to install it

This package has been published in pypi and in order to install it you have

pip install simio-lisa

How to use it

Exporting Output Tables

import os
from simio_lisa.load_simio_project import load_output_tables


if __name__ == '__main__':
    env_project_path = "path to project"
    env_project_file = "name of .simproj file"
    env_model_name = "name of the model containing the output file (usually Model)"
    env_export_dir = "directory where output tables are going to be saved"
    os.mkdir(env_export_dir)
    output_tables = OutputTables(path_to_project=env_project_path,
                                 model_file_name=env_project_file,
                                 model_name=env_model_name)
    output_tables.load_output_tables()
    for table_name, table_df in output_tables.tables.items():
        print(os.path.join(env_export_dir, f'{table_name}.csv'))
        try:
            for col_name, col_type in table_df.dtypes.items():
                if col_type.name == 'datetime64[ns]':
                    table_df[col_name] = table_df[col_name].dt.strftime('%d-%m-%Y %X')
            table_df.to_csv(os.path.join(env_export_dir, f'{table_name}.csv'), index=False, decimal='.')
        except AttributeError:
            print("This was empty")

Exporting Experiments

import os
from simio_lisa.load_simio_project import load_experiment_results


if __name__ == '__main__':
    env_project_path = "path to project"
    env_project_file = "name of .simproj file"
    env_model_name = "name of the model containing the output file (usually Model)"
    experiments_df = load_experiment_results(project_path=env_project_path,
                                             project_filename=env_project_file,
                                             model_name=env_model_name,
                                             agg_function=np.mean)

Plotting Data from tables

Different classes are defined for different kinds of plot. Their parent class is SimioPlotter, and it wants as an input a dictionary with all the tables (e.g. the attribute tables of an object of the class OutputTables). Other possible inputs can be x_axis, y_axis, time_axis, legend_col, object_groups_dict. Each child class must cointain a plot() method. The child classes are: SimioTimeSeries (plot time series), SimioBarPie (bar plots and pie charts), SimioBox (box plot), SimioStackedBars (stacked bars plot).

Examples for the

Initialize OutputTables class object

output_tables = OutputTables(path_to_project,
                             model_file_name,
                             model_name)
output_tables.load_output_tables()

Plot time series comparing different columns of the same table

y_axis = 'Utilization'
time_axis = 'DateTime'
simio_time_series_plotter = SimioTimeSeries(
                  output_tables=output_tables.tables,
                  logger_level = logging.INFO,
                  y_axis= y_axis,
                  time_axis=time_axis)

simio_time_series_plotter.plot(tables='OutputObjectUtilization', kind='time_series_columns')

Plot time series comparing same column from different tables (name of tables as legend)

y_axis = 'Count'
time_axis = 'StatusDate'
simio_time_series_plotter = SimioTimeSeries(
                  output_tables=output_tables.tables,
                  logger_level = logging.INFO,
                  y_axis= y_axis,
                  time_axis=time_axis)

simio_time_series_plotter.plot(tables=['OutputStatus5A', 'OutputStatus5B', 'OutputStatus6'], kind='time_series_tables')

Plot bars or pie charts, distinguishing plots via object_groups_dict dictionary

x_axis = 'ObjectId'
y_axis = 'Utilization'
object_groups_dict = {'Shuttles': ['DropOffShuttle[1]', 'PickUpShuttle[1]'],
                      'Retorts': ['Retort1', 'Retort2', 'Retort3', 'Retort4',
                                  'Retort5', 'Retort6', 'Retort7', 'Retort8',
                                  'Retort9', 'Retort10']
                      }
simio_obj_util_plotter = SimioBarPie(
                  output_tables=output_tables.tables,
                  logger_level = logging.INFO,
                  x_axis = x_axis,
                  y_axis = y_axis,
                  objects_dict = object_groups_dict)

simio_obj_util_plotter.plot(tables='OutputObjectUtilization', kind='bars_plot')
simio_obj_util_plotter.plot(tables='OutputObjectUtilization', kind='pie_plot')

Plot bars along time, distinguishing plots via object_groups_dict dictionary (each key should contain all the objects to be compared together)

x_axis = 'ObjectId'
y_axis = 'Utilization'
time_axis = 'DateTime'
object_groups_dict = {'Shuttles': ['DropOffShuttle[1]', 'PickUpShuttle[1]'],
                      'Retorts': ['Retort1', 'Retort2', 'Retort3', 'Retort4',
                                  'Retort5', 'Retort6', 'Retort7', 'Retort8',
                                  'Retort9', 'Retort10']
                      }
simio_obj_util_plotter = SimioBarPie(
                  output_tables=output_tables.tables,
                  logger_level = logging.INFO,
                  x_axis = x_axis,
                  y_axis = y_axis,
                  time_axis = time_axis,
                  objects_dict = object_groups_dict)

simio_obj_util_plotter.plot(tables='OutputObjectUtilization', kind='bars_time_series_plot')

Box plot

x_axis = 'ProcessName'
y_axis = 'ProductTimeInSystem'
simio_tis_plotter = SimioBox(
    output_tables=output_tables.tables,
    logger_level=logging.INFO,
    y_axis=y_axis,
    x_axis=x_axis)

simio_tis_plotter.plot(tables='OutputProductDeparture', kind='box_plot')

Plot stacked bars, using as a legend the column legend_col

x_axis = 'ObjectID'
y_axis = 'Duration'
legend_col = 'OperationID'
simio_object_processing_plotter = SimioStackedBars(
                  output_tables=output_tables.tables,
                  logger_level = logging.INFO,
                  x_axis = x_axis,
                  y_axis = y_axis,
                  legend_col = legend_col)
simio_object_processing_plotter.plot(tables='ObjectProcessingTable', kind='stacked_bars')