Stages in different columns

[ghost]

Hello and thank you for pySPA!
Is there a way to import the export_to_csv() results back into Python in a way that stages are given in different columns? I've used pd.read_csv('file.csv') and all results are given in a single column.

[ghost]

If multiregional is set to multiregional=True in export_to_csv(), then the region is appended to the sector in brackets. Hence, the string in that column can be cut apart based on the position of the brackets :) ... just a workaround though. Alternatively, a specific string can be inserted in def print_pathway() in the line with if multiregional.

[rrodorr]

The issues seems to be that the csv output has its rows wrapped in quotation marks, rather than just row elements. The raw output looks like this:

"Target sector: 	Cattle farming"
"Sector ID: 	824"
"Number of Regions in input data: 	49"
"Number of sectors in A matrix: 	7987"
"Total number of pathways extracted: 	457"
"Stages analysed: 	8"

while it should look like this:

"Target sector:" 	"Cattle farming"
"Sector ID:" 	"824"
"Number of Regions in input data:" 	"49"
"Number of sectors in A matrix:" 	"7987"
"Total number of pathways extracted:" 	"457"
"Stages analysed:" 	"8"

or e.g. like this (delimiter ;):

Target sector:;Cattle farming
Sector ID:;824
Number of Regions in input data:;49
Number of sectors in A matrix:;7987
Total number of pathways extracted:;457
Stages analysed:;8

If you take a look at the source code, it seems the row elements are deliberately joined for csv export (cf. if to_csv). I don't quite understand why. Maybe to fix some issue that had occured downstream? Granted, there is a varying number of columns in the csv output, but Excel displays this correctly nonetheless. To me, this presents a considerable hurdle for working with the output data. I am now thinking about parsing the csv text and replacing the \t with ; delimiters and dropping the row-wrapping (instead of element-wrapping) quotation marks in the original csv file.

def _generate_spa_title_block(self, flow, to_csv=True or False):
        """
        Creates a title block for the SPA results produced for each flows.
        :param flow: flow being analysed
        :param to_csv: boolean flag referencing whether the data is to be saved to csv file or not
        :return: title block of the SPA results for a particular flow
        """
        try:
            flow_unit = self.flows_dict[flow].unit
        except AttributeError:
            raise AttributeError('The flow %s is not defined' % flow)

        spa_title_block_list = [
            ['flow analysed:', flow],
            ['unit:',
             "%s/%s" % (flow_unit, self.root_node.get_node_attribute(self.sector_definition_dict, 'Unit'))],
            ['thresholds:', str(self.thresholds_dict[flow])],
            ['direct:', str(self.root_node.direct_intensities[flow]), "%s/%s" %
             (flow_unit, self.root_node.get_node_attribute(self.sector_definition_dict, 'Unit'))],
            ['total:', str(self.root_node.total_intensities[flow]), "%s/%s" %
             (flow_unit, self.root_node.get_node_attribute(self.sector_definition_dict, 'Unit'))],
            ['% of total covered by SPA:', "{:.2%}".format(self.get_coverage_of(flow)),
             'Note: Value may differ from sum of percentages in the table due to rounding']
        ]
        if to_csv:
            new_spa_title_block_list = []
            for row in spa_title_block_list:
                new_row = '\t'.join(row)
                new_spa_title_block_list.append(new_row)
            return new_spa_title_block_list
        else:
            return spa_title_block_list

[Geoffrey-Guest]

thanks for the package guys! I just applied it to ecoinvent.

Here's what I did to create a column-wise df:

spaResults = pd.read_csv("spa_results.csv")
spaSummary = spaResults[0:16]
spaTable_header = spaResults.loc[17].tolist()[0].split("\t")
spaPaths = spaResults[18:].reset_index(drop=True)
spaPaths = spaPaths[spaPaths.columns[0]].tolist()

spaTable = pd.DataFrame([(x.split('\t')+['']*(18-len(x.split('\t')))) for x in spaPaths], columns=spaTable_header)

spaTable