statistics.md

Object statistics

You can get statistical data about objects in a csv file using the Totals class.

from pipeline_csv.oegiv import File
from pipeline_csv.csvfile.statistics.totals import Totals

csv_file = File.from_file('DefTable.csv', 1400)
totals = Totals()
warns = []
totals.fill(csv_file, warns)
assert not warns

The statistical data contains the start and length marks of the section, as well as a list of markers.

assert totals.start == 0
assert totals.length == 426625
assert len(totals.markers) == 5

The csv file contains 14 linear objects, which are located on 13 pipes.

assert totals.liners.number == 14
assert totals.liners.tubes_all() == 13

Total number of pipes in csv file: 41.

assert totals.pipes.number == 41

There are three pipe wall thicknesses in total: 7, 9, and 10 mm.

assert list(sorted(totals.pipes.thick.data.keys())) == [70, 90, 100]

The number of pipes with a wall thickness of 7 mm is 22 units, with a total length of 232232 mm.

pipes_7_mm = totals.pipes.thick.data[70]
assert pipes_7_mm.number == 22
assert pipes_7_mm.length == 232232

All pipes in the file are of the same type - straight-seam.

from pipeline_csv.oegiv import TypeDefekt

assert list(totals.pipes.types.data.keys()) == [TypeDefekt.HORIZONTAL]

There are 75 defects in the file.

assert totals.defects.number == 75

By location on the pipe wall, there are 6 defects with an undefined position and 69 external defects.

from pipeline_csv import DefektSide

assert list(sorted(totals.defects.base_wallside.data.keys())) == [DefektSide.UNKNOWN, DefektSide.OUTSIDE]
assert totals.defects.base_wallside.data[DefektSide.UNKNOWN].number == 6
assert totals.defects.base_wallside.data[DefektSide.OUTSIDE].number == 69

There are 6 types of defects.

types = totals.defects.base_types.data
assert len(types) == 6

56 defects of 'corrosion', 6 mechanical damage, etc.

assert types[TypeDefekt.CORROZ].number == 56
assert types[TypeDefekt.MECHANIC].number == 6
assert types[TypeDefekt.DENT].number == 2
assert types[TypeDefekt.GWAN].number == 1
assert types[TypeDefekt.TECHNOLOGY].number == 8
assert types[TypeDefekt.FACTORY].number == 2

Distribution of defects by orientation hours.

assert totals.defects.base_angle_anomalies.hours == {
  0: 6,  # 6 defects in the 12 o'clock sector
  1: 3,  # 3 defects in the 1 o'clock sector
  2: 6,  # etc
  3: 8,
  4: 8,
  5: 7,
  6: 9,
  7: 15,
  8: 9,
  9: 17,
  10: 7,
  11: 8
}

The Totals class also allows you to obtain extended statistics by overriding the statistics classes for pipes and defects.

To obtain defect statistics based on your own requirements, you need to define a class inheriting from the pipeline_csv.csvfile.statistics.defects.Totals base class. In this class, you need to override the __init__ and add_defect methods.

from pipeline_csv.csvfile.statistics.defects import Totals as DefectsTotalsBase, Dents

class DefectsTotals(DefectsTotalsBase):
    """Custom defect totals class."""

    def __init__(self, start, length, markers):
        """Make new defects total object with custom properties."""
        super().__init__(start, length, markers)
        self.dents = Dents(grades=[5, 10])

    def add_defect(self, defect, tube, warns):
        """Add defect to custom statistics."""
        super().add_defect(defect, tube, warns)
        if defect.is_dent:
            self.dents.add_data(defect)

This class uses the Dents library class to collect user-defined statistics grading dent depth.

• up to 5% of the diameter
• from 5% to 10% of the diameter
• more than 10% of the diameter

When creating an instance of the CSV file statistics class, you must pass the name of the user-defined class as a parameter.

from pipeline_csv.csvfile.statistics.totals import Totals

totals = Totals(defects_class=DefectsTotals)
totals.fill(csv_file, None)

After this, you will have access to dent statistics.

There are two dents in total.

assert totals.defects.dents.number == 2

There are no dents deeper than 10% of the diameter.

from pipeline_csv.csvfile.statistics.totals import GRADE_OVER_MAX

assert totals.defects.dents.data[GRADE_OVER_MAX] == 0

There is one dent with a depth of up to 5% and one from 5 to 10%.

assert totals.defects.dents.data[5] == 1
assert totals.defects.dents.data[10] == 1

A dent of up to 5% depth is found on pipe number W6332, and a dent of up to 10% depth is found on pipe number W14736.

assert list(totals.defects.dents.tubes[5].keys()) == ['W6332']
assert list(totals.defects.dents.tubes[10].keys()) == ['W14736']

You can use the following defect statistics library classes:

• DistSingle: distribution by distance
• DistWallside: distribution by distance broken down by position on the pipe wall
• DangerValve: distribution between valves
• Depth: grouping metal loss by specified depth intervals (as a percentage of the pipe wall thickness)
• Dents: grouping dents by specified depth intervals (as a percentage of the pipe diameter)
• Angles: distribution by orientation hours
• PropertyCounter: distribution by values of a specified defect property

You can also define your own defect statistics classes to collect the required data.