def arrange_into_groups()

in src/filter_and_group.py [0:0]

• 33 lines of code
• 14 McCabe index (conditional complexity)

def arrange_into_groups(datasets, group_by, column, column_timing, labels):
        timestamp_groups = []
        datapoint_groups = []
        group_labels = []
        for idx, df in enumerate(datasets): # Loop through all provided log datasets
            if not df.empty:
                if group_by not in df:
                    print("Warning: group_by group " + str(group_by) + " column not in dataset with columns " + str(df.columns))
                elif column not in df:
                    print("Warning: column \"" + str(column) + "\" not in dataset with columns " + str(df.columns))
                elif column_timing not in df:
                    print("Warning: column_timing \"" + str(column_timing) + "\" not in dataset with columns " + str(df.columns))
                else: 
                    # A non-empty df contains both X and Y columns.                    
                    
                    groups = {} # Create a dictionary to hold unique groups
                    if column_timing == "DateTime":
                        print("Case number 1")
                        timing = pd.Series(matplotlib.dates.date2num(df[column_timing]))
                    else:
                        timing = df[column_timing]                        
                    for group, time, datapoint in zip(df[group_by], timing, df[column]):
                        if not group:
                            group = "( " + str(group_by) + " = None )" # None groups should all be put together
                        if group not in groups:
                            # Create a new group for each unique item
                            groups[group] = [[], [], str(labels[idx]) + ": " + str(group)]
                        
                        # add the datapoints and time, based on the grouping
                        groups[group][0].append(time)
                        groups[group][1].append(datapoint)

                    # Sort keys so groups print in the same order between files
                    keys = list(groups.keys())
                    keys.sort()

                    for key in keys:
                        timestamp_groups.append(pd.Series(groups[key][0]))
                        datapoint_groups.append(pd.Series(groups[key][1]))
                        group_labels.append(groups[key][2])
                        
        return timestamp_groups, datapoint_groups, group_labels