Source code for flopy.modpath.mp7particledata

"""
mp7particledata module. Contains the ParticleData, CellDataType,
    FaceDataType, and NodeParticleData classes.


"""

import numpy as np

from ..utils.recarray_utils import create_empty_recarray


[docs]class ParticleData: """ Class to create the most basic particle data type (starting location input style 1). Input style 1 is the most general input style and provides the most flexibility in customizing starting locations. Parameters ---------- partlocs : list/tuple of int, list/tuple of list/tuple, or np.ndarray Particle locations (zero-based) that are either layer, row, column locations or nodes. structured : bool Boolean defining if a structured (True) or unstructured particle recarray will be created (default is True). particleids : list, tuple, or np.ndarray Particle ids for the defined particle locations. If particleids is None, MODPATH 7 will define the particle ids to each particle location. If particleids is provided a particle id must be provided for each partloc (default is None). localx : float, list, tuple, or np.ndarray Local x-location of the particle in the cell. If a single value is provided all particles will have the same localx position. If a list, tuple, or np.ndarray is provided a localx position must be provided for each partloc. If localx is None, a value of 0.5 (center of the cell) will be used (default is None). localy : float, list, tuple, or np.ndarray Local y-location of the particle in the cell. If a single value is provided all particles will have the same localy position. If a list, tuple, or np.ndarray is provided a localy position must be provided for each partloc. If localy is None, a value of 0.5 (center of the cell) will be used (default is None). localz : float, list, tuple, or np.ndarray Local z-location of the particle in the cell. If a single value is provided all particles will have the same localz position. If a list, tuple, or np.ndarray is provided a localz position must be provided for each partloc. If localy is None, a value of 0.5 (center of the cell) will be used (default is None). timeoffset : float, list, tuple, or np.ndarray Timeoffset of the particle relative to the release time. If a single value is provided all particles will have the same timeoffset. If a list, tuple, or np.ndarray is provided a timeoffset must be provided for each partloc. If timeoffset is None, a value of 0. (equal to the release time) will be used (default is None). drape : int, list, tuple, or np.ndarray Drape indicates how particles are treated when starting locations are specified for cells that are dry. If drape is 0, Particles are placed in the specified cell. If the cell is dry at the time of release, the status of the particle is set to unreleased and removed from the simulation. If drape is 1, particles are placed in the upper most active grid cell directly beneath the specified layer, row, column or node location. If a single value is provided all particles will have the same drape value. If a list, tuple, or np.ndarray is provided a drape value must be provided for each partloc. If drape is None, a value of 0 will be used (default is None). Examples -------- >>> import flopy >>> locs = [(0, 0, 0), (1, 0, 0), (2, 0, 0)] >>> pd = flopy.modpath.ParticleData(locs, structured=True, drape=0, ... localx=0.5, localy=0.5, localz=1) """ def __init__( self, partlocs=None, structured=False, particleids=None, localx=None, localy=None, localz=None, timeoffset=None, drape=None, ): """ Class constructor """ self.name = "ParticleData" if structured: locationstyle = 1 else: locationstyle = 2 if partlocs is None: if structured: partlocs = [(0, 0, 0)] else: partlocs = [(0,)] # create dtype dtype = [] if structured: dtype.append(("k", np.int32)) dtype.append(("i", np.int32)) dtype.append(("j", np.int32)) else: dtype.append(("node", np.int32)) dtype = np.dtype(dtype) if isinstance(partlocs, (list, tuple)): # determine if the list or tuple contains lists or tuples alllsttup = all(isinstance(el, (list, tuple)) for el in partlocs) if structured: if alllsttup: alllen3 = all(len(el) == 3 for el in partlocs) if not alllen3: raise ValueError( "{}: all partlocs entries must have 3 items for " "structured particle data".format(self.name) ) else: raise ValueError( "{}: partlocs list or tuple " "for structured particle data should " "contain list or tuple entries".format(self.name) ) else: allint = all( isinstance(el, (int, np.int32, np.int64)) for el in partlocs ) # convert to a list of tuples if allint: t = [] for el in partlocs: t.append((el,)) partlocs = t alllsttup = all( isinstance(el, (list, tuple)) for el in partlocs ) if alllsttup: alllen1 = all(len(el) == 1 for el in partlocs) if not alllen1: raise ValueError( "{}: all entries of partlocs must have 1 items " "for unstructured particle data".format(self.name) ) else: raise ValueError( "{}: partlocs list or tuple for unstructured particle " "data should contain integers or a list or tuple with " "one entry".format(self.name) ) # convert partlocs composed of a lists/tuples of lists/tuples # to a numpy array partlocs = np.array(partlocs, dtype=dtype) elif isinstance(partlocs, np.ndarray): dtypein = partlocs.dtype if dtypein != dtype: partlocs = np.array(partlocs, dtype=dtype) else: raise ValueError( f"{self.name}: partlocs must be a list or tuple with lists or tuples" ) # localx if localx is None: localx = 0.5 else: if isinstance(localx, (float, int)): localx = np.ones(partlocs.shape[0], dtype=np.float32) * localx elif isinstance(localx, (list, tuple)): localx = np.array(localx, dtype=np.float32) if isinstance(localx, np.ndarray): if localx.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of localx ({}) is not equal to the shape " "of partlocs ({}).".format( self.name, localx.shape[0], partlocs.shape[0] ) ) # localy if localy is None: localy = 0.5 else: if isinstance(localy, (float, int)): localy = np.ones(partlocs.shape[0], dtype=np.float32) * localy elif isinstance(localy, (list, tuple)): localy = np.array(localy, dtype=np.float32) if isinstance(localy, np.ndarray): if localy.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of localy ({}) is not equal to the shape " "of partlocs ({}).".format( self.name, localy.shape[0], partlocs.shape[0] ) ) # localz if localz is None: localz = 0.5 else: if isinstance(localz, (float, int)): localz = np.ones(partlocs.shape[0], dtype=np.float32) * localz elif isinstance(localz, (list, tuple)): localz = np.array(localz, dtype=np.float32) if isinstance(localz, np.ndarray): if localz.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of localz ({}) is not equal to the shape " "of partlocs ({}).".format( self.name, localz.shape[0], partlocs.shape[0] ) ) # timeoffset if timeoffset is None: timeoffset = 0.0 else: if isinstance(timeoffset, (float, int)): timeoffset = ( np.ones(partlocs.shape[0], dtype=np.float32) * timeoffset ) elif isinstance(timeoffset, (list, tuple)): timeoffset = np.array(timeoffset, dtype=np.float32) if isinstance(timeoffset, np.ndarray): if timeoffset.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of timeoffset ({}) is not equal to the " "shape of partlocs ({}).".format( self.name, timeoffset.shape[0], partlocs.shape[0] ) ) # drape if drape is None: drape = 0 else: if isinstance(drape, (float, int)): drape = np.ones(partlocs.shape[0], dtype=np.float32) * drape elif isinstance(drape, (list, tuple)): drape = np.array(drape, dtype=np.int32) if isinstance(drape, np.ndarray): if drape.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of drape ({}) is not equal to the shape " "of partlocs ({}).".format( self.name, drape.shape[0], partlocs.shape[0] ) ) # particleids if particleids is None: particleid = False particleidoption = 0 else: particleid = True particleidoption = 1 if isinstance(particleids, (int, float)): raise TypeError( "{}:A particleid must be provided for each partloc " "as a list/tuple/np.ndarray of size {}. " "A single particleid has been provided.".format( self.name, partlocs.shape[0] ) ) elif isinstance(particleids, (list, tuple)): particleids = np.array(particleids, dtype=np.int32) if isinstance(particleids, np.ndarray): if particleids.shape[0] != partlocs.shape[0]: raise ValueError( "{}:shape of particleids ({}) is not equal to the " "shape of partlocs ({}).".format( self.name, particleids.shape[0], partlocs.shape[0] ) ) # create empty particle ncells = partlocs.shape[0] self.dtype = self._get_dtype(structured, particleid) particledata = create_empty_recarray( ncells, self.dtype, default_value=0 ) # fill particle if structured: particledata["k"] = partlocs["k"] particledata["i"] = partlocs["i"] particledata["j"] = partlocs["j"] else: particledata["node"] = partlocs["node"] particledata["localx"] = localx particledata["localy"] = localy particledata["localz"] = localz particledata["timeoffset"] = timeoffset particledata["drape"] = drape if particleid: particledata["id"] = particleids self.particlecount = particledata.shape[0] self.particleidoption = particleidoption self.locationstyle = locationstyle self.particledata = particledata return
[docs] def write(self, f=None): """ Parameters ---------- f : fileobject Fileobject that is open with write access Returns ------- """ # validate that a valid file object was passed if not hasattr(f, "write"): raise ValueError( "{}: cannot write data for template without passing a valid " "file object ({}) open for writing".format(self.name, f) ) # particle data item 4 and 5 d = np.recarray.copy(self.particledata) lnames = [name.lower() for name in d.dtype.names] # Add one to the kij and node indices for idx in ( "k", "i", "j", "node", ): if idx in lnames: d[idx] += 1 # Add one to the particle id if required if self.particleidoption == 0 and "id" in lnames: d["id"] += 1 # write the particle data fmt = self._fmt_string + "\n" for v in d: f.write(fmt.format(*v)) return
def _get_dtype(self, structured, particleid): """ define the dtype for a structured or unstructured particledata recarray. Optionally, include a particleid column in the dtype. Parameters ---------- structured : bool Boolean defining if a structured (True) or unstructured particle dtype will be created. particleid : bool Boolean defining if the dtype will include a particle id column. Returns ------- dtype : numpy dtype Examples -------- >>> import flopy.modpath as fmp >>> dtype = fmp.ParticleGroup.get_particledata_dtype(structured=True, ... particleid=True) """ dtype = [] if particleid: dtype.append(("id", np.int32)) if structured: dtype.append(("k", np.int32)) dtype.append(("i", np.int32)) dtype.append(("j", np.int32)) else: dtype.append(("node", np.int32)) dtype.append(("localx", np.float32)) dtype.append(("localy", np.float32)) dtype.append(("localz", np.float32)) dtype.append(("timeoffset", np.float32)) dtype.append(("drape", np.int32)) return np.dtype(dtype) @property def _fmt_string(self): """ Returns a python-style fmt string to write particle data that corresponds to the dtype Parameters ---------- Returns ------- fmt : str python format string with space delimited entries """ fmts = [] for field in self.particledata.dtype.descr: vtype = field[1][1].lower() if vtype == "i" or vtype == "b": fmts.append("{:9d}") elif vtype == "f": if field[1][2] == 8: fmts.append("{:23.16g}") else: fmts.append("{:15.7g}") elif vtype == "o": fmts.append("{:9s}") elif vtype == "s": raise TypeError( "Particles.fmt_string error: 'str' type found in dtype. " "This gives unpredictable results when recarray to file - " "change to 'object' type" ) else: raise TypeError( f"MfList.fmt_string error: unknown vtype in field: {field}" ) return " " + " ".join(fmts)
[docs]class FaceDataType: """ Face data type class to create a MODPATH 7 particle location template for input style 2, 3, and 4 on cell faces (templatesubdivisiontype = 2). Parameters ---------- drape : int Drape indicates how particles are treated when starting locations are specified for cells that are dry. If drape is 0, Particles are placed in the specified cell. If the cell is dry at the time of release, the status of the particle is set to unreleased and removed from the simulation. If drape is 1, particles are placed in the upper most active grid cell directly beneath the specified layer, row, column or node location (default is 0). verticaldivisions1 : int The number of vertical subdivisions that define the two-dimensional array of particles on cell face 1 (default is 3). horizontaldivisions1 : int The number of horizontal subdivisions that define the two-dimensional array of particles on cell face 1 (default is 3). verticaldivisions2 : int The number of vertical subdivisions that define the two-dimensional array of particles on cell face 2 (default is 3). horizontaldivisions2 : int The number of horizontal subdivisions that define the two-dimensional array of particles on cell face 2 (default is 3). verticaldivisions3 : int The number of vertical subdivisions that define the two-dimensional array of particles on cell face 3 (default is 3). horizontaldivisions3 : int The number of horizontal subdivisions that define the two-dimensional array of particles on cell face 3 (default is 3). verticaldivisions4 : int The number of vertical subdivisions that define the two-dimensional array of particles on cell face 4 (default is 3). horizontaldivisions4 : int The number of horizontal subdivisions that define the two-dimensional array of particles on cell face 4 (default is 3). rowdivisions5 : int The number of row subdivisions that define the two-dimensional array of particles on the bottom cell face (face 5) (default is 3). columndivisions5 : int The number of column subdivisions that define the two-dimensional array of particles on the bottom cell face (face 5) (default is 3). rowdivisions6 : int The number of row subdivisions that define the two-dimensional array of particles on the top cell face (face 6) (default is 3). columndivisions6 : int The number of column subdivisions that define the two-dimensional array of particles on the top cell face (face 6) (default is 3). Examples -------- >>> import flopy >>> fd = flopy.modpath.FaceDataType() """ def __init__( self, drape=0, verticaldivisions1=3, horizontaldivisions1=3, verticaldivisions2=3, horizontaldivisions2=3, verticaldivisions3=3, horizontaldivisions3=3, verticaldivisions4=3, horizontaldivisions4=3, rowdivisions5=3, columndivisions5=3, rowdivisions6=3, columndivisions6=3, ): """ Class constructor """ self.name = "FaceDataType" # assign attributes self.templatesubdivisiontype = 1 self.drape = drape self.verticaldivisions1 = verticaldivisions1 self.horizontaldivisions1 = horizontaldivisions1 self.verticaldivisions2 = verticaldivisions2 self.horizontaldivisions2 = horizontaldivisions2 self.verticaldivisions3 = verticaldivisions3 self.horizontaldivisions3 = horizontaldivisions3 self.verticaldivisions4 = verticaldivisions4 self.horizontaldivisions4 = horizontaldivisions4 self.rowdivisions5 = rowdivisions5 self.columndivisions5 = columndivisions5 self.rowdivisions6 = rowdivisions6 self.columndivisions6 = columndivisions6 return
[docs] def write(self, f=None): """ Parameters ---------- f : fileobject Fileobject that is open with write access Returns ------- """ # validate that a valid file object was passed if not hasattr(f, "write"): raise ValueError( "{}: cannot write data for template " "without passing a valid file object ({}) " "open for writing".format(self.name, f) ) # item 4 fmt = 12 * " {}" + "\n" line = fmt.format( self.verticaldivisions1, self.horizontaldivisions1, self.verticaldivisions2, self.horizontaldivisions2, self.verticaldivisions3, self.horizontaldivisions3, self.verticaldivisions4, self.horizontaldivisions4, self.rowdivisions5, self.columndivisions5, self.rowdivisions6, self.columndivisions6, ) f.write(line) return
[docs]class CellDataType: """ Cell data type class to create a MODPATH 7 particle location template for input style 2, 3, and 4 in cells (templatesubdivisiontype = 2). Parameters ---------- drape : int Drape indicates how particles are treated when starting locations are specified for cells that are dry. If drape is 0, Particles are placed in the specified cell. If the cell is dry at the time of release, the status of the particle is set to unreleased and removed from the simulation. If drape is 1, particles are placed in the upper most active grid cell directly beneath the specified layer, row, column or node location (default is 0). columncelldivisions : int Number of particles in a cell in the column (x-coordinate) direction (default is 3). rowcelldivisions : int Number of particles in a cell in the row (y-coordinate) direction (default is 3). layercelldivisions : int Number of particles in a cell in the layer (z-coordinate) direction (default is 3). Examples -------- >>> import flopy >>> cd = flopy.modpath.CellDataType() """ def __init__( self, drape=0, columncelldivisions=3, rowcelldivisions=3, layercelldivisions=3, ): """ Class constructor """ self.name = "CellDataType" # assign attributes self.templatesubdivisiontype = 2 self.drape = drape self.columncelldivisions = columncelldivisions self.rowcelldivisions = rowcelldivisions self.layercelldivisions = layercelldivisions return
[docs] def write(self, f=None): """ Parameters ---------- f : fileobject Fileobject that is open with write access Returns ------- """ # validate that a valid file object was passed if not hasattr(f, "write"): raise ValueError( "{}: cannot write data for template " "without passing a valid file object ({}) " "open for writing".format(self.name, f) ) # item 5 fmt = " {} {} {}\n" line = fmt.format( self.columncelldivisions, self.rowcelldivisions, self.layercelldivisions, ) f.write(line) return
[docs]class LRCParticleData: """ Layer, row, column particle data template class to create MODPATH 7 particle location input style 2 on cell faces (templatesubdivisiontype = 1) and/or in cells (templatesubdivisiontype = 2). Particle locations for this template are specified by layer, row, column regions. Parameters ---------- subdivisiondata : FaceDataType, CellDataType or list of FaceDataType and/or CellDataType types FaceDataType, CellDataType, or a list of FaceDataType and/or CellDataTypes that are used to create one or more particle templates in a particle group. If subdivisiondata is None, a default CellDataType with 27 particles per cell will be created (default is None). lrcregions : list of lists tuples or np.ndarrays Layer, row, column (zero-based) regions with particles created using the specified template parameters. A region is defined as a list/tuple of minlayer, minrow, mincolumn, maxlayer, maxrow, maxcolumn values. If subdivisiondata is a list, a list/tuple or array of layer, row, column regions with the same length as subdivision data must be provided. If lrcregions is None, particles will be placed in the first model cell (default is None). Examples -------- >>> import flopy >>> pg = flopy.modpath.LRCParticleData(lrcregions=[0, 0, 0, 3, 10, 10]) """ def __init__(self, subdivisiondata=None, lrcregions=None): """ Class constructor """ self.name = "LRCParticleData" if subdivisiondata is None: subdivisiondata = CellDataType() if lrcregions is None: lrcregions = [[0, 0, 0, 0, 0, 0]] if isinstance(subdivisiondata, (CellDataType, FaceDataType)): subdivisiondata = [subdivisiondata] for idx, fd in enumerate(subdivisiondata): if not isinstance(fd, (CellDataType, FaceDataType)): raise TypeError( "{}: facedata item {} is of type {} instead of an " "instance of CellDataType or FaceDataType".format( self.name, idx, type(fd) ) ) # validate lrcregions data if isinstance(lrcregions, (list, tuple)): # determine if the list or tuple contains lists or tuples alllsttup = all( isinstance(el, (list, tuple, np.ndarray)) for el in lrcregions ) if not alllsttup: raise TypeError( "{}: lrcregions should be " "a list with lists, tuples, or arrays".format(self.name) ) t = [] for lrcregion in lrcregions: t.append(np.array(lrcregion, dtype=np.int32)) lrcregions = t else: raise TypeError( "{}: lrcregions should be a list of lists, tuples, or arrays " "not a {}.".format(self.name, type(lrcregions)) ) # validate size of nodes relative to subdivisiondata shape = len(subdivisiondata) if len(lrcregions) != shape: raise ValueError( "{}: lrcregions data must have {} rows but a total of {} rows " "were provided.".format(self.name, shape, lrcregions.shape[0]) ) # validate that there are 6 columns in each lrcregions entry for idx, lrcregion in enumerate(lrcregions): shapel = lrcregion.shape if len(shapel) == 1: lrcregions[idx] = lrcregion.reshape(1, shapel) shapel = lrcregion[idx].shape if shapel[1] != 6: raise ValueError( "{}: Each lrcregions entry must " "have 6 columns passed lrcregions has " "{} columns".format(self.name, shapel[1]) ) # totalcellregioncount = 0 for lrcregion in lrcregions: totalcellregioncount += lrcregion.shape[0] # assign attributes self.particletemplatecount = shape self.totalcellregioncount = totalcellregioncount self.subdivisiondata = subdivisiondata self.lrcregions = lrcregions return
[docs] def write(self, f=None): """ Parameters ---------- f : fileobject Fileobject that is open with write access Returns ------- """ # validate that a valid file object was passed if not hasattr(f, "write"): raise ValueError( "{}: cannot write data for template " "without passing a valid file object ({}) " "open for writing".format(self.name, f) ) # item 2 f.write(f"{self.particletemplatecount} {self.totalcellregioncount}\n") for sd, lrcregion in zip(self.subdivisiondata, self.lrcregions): # item 3 f.write( f"{sd.templatesubdivisiontype} {lrcregion.shape[0]} {sd.drape}\n" ) # item 4 or 5 sd.write(f) # item 6 for row in lrcregion: line = "" for lrc in row: line += f"{lrc + 1} " line += "\n" f.write(line) return
[docs]class NodeParticleData: """ Node particle data template class to create MODPATH 7 particle location input style 3 on cell faces (templatesubdivisiontype = 1) and/or in cells (templatesubdivisiontype = 2). Particle locations for this template are specified by nodes. Parameters ---------- subdivisiondata : FaceDataType, CellDataType or list of FaceDataType and/or CellDataType types FaceDataType, CellDataType, or a list of FaceDataType and/or CellDataTypes that are used to create one or more particle templates in a particle group. If subdivisiondata is None, a default CellDataType with 27 particles per cell will be created (default is None). nodes : int, list of ints, tuple of ints, or np.ndarray Nodes (zero-based) with particles created using the specified template parameters. If subdivisiondata is a list, a list of nodes with the same length as subdivision data must be provided. If nodes is None, particles will be placed in the first model cell (default is None). Examples -------- >>> import flopy >>> pg = flopy.modpath.NodeParticleData(nodes=[100, 101]) """ def __init__(self, subdivisiondata=None, nodes=None): """ Class constructor """ self.name = "NodeParticleData" if subdivisiondata is None: subdivisiondata = CellDataType() if nodes is None: nodes = 0 if isinstance(subdivisiondata, (CellDataType, FaceDataType)): subdivisiondata = [subdivisiondata] if isinstance(nodes, (int, np.int32, np.int64)): nodes = [(nodes,)] elif isinstance(nodes, (float, np.float32, np.float64)): raise TypeError( "{}: nodes is of type {} but must be an int if a " "single value is passed".format(self.name, type(nodes)) ) for idx, fd in enumerate(subdivisiondata): if not isinstance(fd, (CellDataType, FaceDataType)): raise TypeError( "{}: facedata item {} is of type {} instead of an " "instance of CellDataType or FaceDataType".format( self.name, idx, type(fd) ) ) # validate nodes data if isinstance(nodes, np.ndarray): if len(nodes.shape) == 1: nodes = nodes.reshape(1, nodes.shape[0]) # convert to a list of numpy arrays t = [] for idx in range(nodes.shape[0]): t.append(np.array(nodes[idx, :], dtype=np.int32)) nodes = t elif isinstance(nodes, (list, tuple)): # convert a single list/tuple to a list of tuples if only one # entry in subdivisiondata if len(subdivisiondata) == 1: if len(nodes) > 1: nodes = [tuple(nodes)] # determine if the list or tuple contains lists or tuples alllsttup = all( isinstance(el, (list, tuple, np.ndarray)) for el in nodes ) if not alllsttup: raise TypeError( "{}: nodes should be " "a list or tuple with lists or tuple if a single " "int or numpy array is not provided".format(self.name) ) t = [] for idx in range(len(nodes)): t.append(np.array(nodes[idx], dtype=np.int32)) nodes = t else: raise TypeError( "{}: nodes should be a single integer, a numpy array, or a " "list/tuple or lists/tuples.".format(self.name) ) # validate size of nodes relative to subdivisiondata shape = len(subdivisiondata) if len(nodes) != shape: raise ValueError( "{}: node data must have {} rows but a total of {} rows were " "provided.".format(self.name, shape, nodes.shape[0]) ) totalcellcount = 0 for t in nodes: totalcellcount += t.shape[0] # assign attributes self.particletemplatecount = shape self.totalcellcount = totalcellcount self.subdivisiondata = subdivisiondata self.nodedata = nodes return
[docs] def write(self, f=None): """ Parameters ---------- f : fileobject Fileobject that is open with write access Returns ------- """ # validate that a valid file object was passed if not hasattr(f, "write"): raise ValueError( "{}: cannot write data for template " "without passing a valid file object ({}) " "open for writing".format(self.name, f) ) # item 2 f.write(f"{self.particletemplatecount} {self.totalcellcount}\n") for sd, nodes in zip(self.subdivisiondata, self.nodedata): # item 3 f.write( f"{sd.templatesubdivisiontype} {nodes.shape[0]} {sd.drape}\n" ) # item 4 or 5 sd.write(f) # item 6 line = "" for idx, node in enumerate(nodes): line += f" {node + 1}" lineend = False if idx > 0: if idx % 10 == 0 or idx == nodes.shape[0] - 1: lineend = True if lineend: line += "\n" f.write(line) return