fonsim.data package

Submodules

fonsim.data.curve module

Class Curve

2020, September 1

class fonsim.data.curve.Curve(data, key_x, key_f, convert_to_base_si=False, autocorrect=False, **interpolation_opts)[source]

Bases: object

Class to ease working with pv- and pn-curves and similar curves

Parameters: data – filepath to CSV file or DataSeries-like object

autocorrect(arg)[source]

TODO implement autocorrect here I made this a separate function such that child classes can modify the data as they want before they use the autocorrect functionality.

Return None

fdf(x)[source]

Readout f(x)

Parameters: x – x
Returns: f, df

fonsim.data.dataseries module

Class DataSeries

2020, September 1

class fonsim.data.dataseries.DataSeries(filename, bytestring=None)[source]

Bases: object

Class to load and hold tabular data from CSV file. Numerical data is stored in numpy arrays as floats. Labels and units are stored in Python lists

Parameters

filename – path to file to read or, if bytestring given, filetype
bytestring – bytestring with file data

load_data(filename, bytestring=None)[source]

Load in data. Provide a filename or byte string. If providing a byte string, provide the filetype extension (e.g. .csv) to the filename argument such that the formatting of the bytestring can be determined.

Parameters

filename – path to file to read or, if bytestring given, filetype
bytestring – bytestring with file data

Returns

None

fonsim.data.interpolate module

Function interpolate_fdf

2020, September 4

fonsim.data.interpolate.interpolate_fdf(x, xs, ys, extrapolate=False, extrapolate_derivative=False, method='linear')[source]

Quickly interpolate a dataseries and return both interpolated value and its derivative.

Arrays xs, ys can be Numpy arrays, yet any object that allows indexing can be used, such as Python lists.

Method: linear or quadratic Search: bisection

TODO

Document pchip method.
Document extrapolation options. The current naming is not great, as choosing ‘False’ for extrapolation results in a zero-order extrapolation. Better to use numbers, e.g. ‘-1’ for no extrapolation (throw error), ‘0’ for zero-order extrapolation and ‘1’ for linear extrapolation?
Shorten parameters extrapolate and extrapolate_derivative?

Parameters

x – x value to interpolate at
xs – x dataseries
ys – y dataseries
extrapolate – True or False
extrapolate_derivative – True or False
method – ‘linear’ or ‘quadratic’

Returns

f and df/dx, both evaluated at x

Note

Given that this function is used very often, it is written with the eye on fast execution rather than modularity and good looks.

fonsim.data.pvcurve module

Class PVCurve

2020, September 4

class fonsim.data.pvcurve.PVCurve(data, pressure_reference='relative', autocorrect=False, **interpolation_opts)[source]

Bases: Curve

Class to ease working with pv-curves

Warning: original data in DataSeries object may be modified by this function. Take a deepcopy if modification undesirable.

The autocorrect functionality provides a little tool to correct measurement data. Parameter autocorrect should be a tuple of length two (respectively volume and pressure) or a scalar. The elements of this tuple, or the scalar, can be:

False: No correction applied.
True: Default correction applied. For volume, the volume at index 0 will equal zero. For pressure, the pressure at index 0 will equal standard atmospheric pressure.
A scalar value: Default correction is applied whereafter an offset with the given value is applied. Units are m^3 for volume and Pa for pressure.
A function: The function is applied to the value series. The function should take the value series as argument and should return the new value series.

Note: the pressure_reference parameter looses its effect when autocorrect is applied to pressure.

Note: The volume data sequence should be increasing or decreasing, otherwise the interpolation function will not work.

Example:

import fonsim

# Create PVCurve object
curve = fonsim.data.pvcurve.PVCurve('mypvcurvefile.csv',
                pressure_reference='relative', autocorrect=True)

# Readout the absolute pressure and its derivative to volume
# at volume 3.8e-6 m^3 (= 3.8 ml)
p, dp_dv = PVCurve.fdf_volume(3.8e-6)

TODO Discuss format of CSV file.

Parameters

data – filepath to CSV file or DataSeries-like object
pressure_reference – “relative” or “absolute”
autocorrect – see description
interpolation_opts – kwargs for interpolation function

fdf_volume(volume)[source]

Readout the pressure for a given volume

Parameters: volume – volume in [m3]
Returns: f, df

get_initial_volume(p0)[source]

Get the volume of the first datapoint on the curve that approaches the provided pressure value the closest

TODO what is this function used for?

Parameters: p0 – pressure at which to find the first matching volume
Returns: first closest matching volume

property p[source]

property v[source]

fonsim.data.writeout module

Function writeout_simulation

2020, September 9

class fonsim.data.writeout.Bank[source]

Bases: object

add(obj)[source]: Add an object. Object does not have to be hashable. :param obj: object :return: index of object, integer

indices()[source]: Return all indices pointing to objects. :return: all indices, Python range

fonsim.data.writeout.writeout_simulation(filename, simulation)[source]

Write out simulation data in components to a file. Supported formats: JSON. Format follows from filename extension.

Parameters

filename – string with filepath
simulation – Simulation-like object

Returns

None

===

JSON specification:

{
  "scheme": <string>,
  "general": {
    "date": <timestamp>,
    "hostname": <computer name",
    "version": <version>
  },
  "simulation": {
    "system": {
      "label": <string>,
      "nb components": <integer>
    },
    "solver": {
      "name": <string>,
    },
    "time": <key in "data"
  },
  "components": {
    <component_label>: {
      "terminals": {
        <terminal_label>: {
          "over": {
            <over_label>: <key in "data">
          },
          through": {
            <through_label>: <key in "data">
          }
        },
        ...
      }
      "states": {
        <state_label>: <key in "data">,
        ...
      },
      "time": <key in "data">
    },
    ...
  },
  "data": {
    <key>: <list with numbers>,
    ...
  },
}

Module contents

2020, September 18