__functions__.py

# -*- coding: utf-8 -*-
"""
Created on Thu Oct 22 12:17:13 2020

@author: modelpc
"""
import pandas as pd
import numpy as np
import shapefile as shp
import array
import os.path


def read_mtx(mtxfile):
    '''
    Read a binary mtx-file (skimTijd and skimAfstand)
    '''
    mtxData = array.array('f')  # f for float
    mtxData.fromfile(
        open(mtxfile, 'rb'),
        os.path.getsize(mtxfile) // mtxData.itemsize)

    # The number of zones is in the first byte
    mtxData = np.array(mtxData, dtype=float)[1:]

    return mtxData


def write_mtx(filename, mat, aantalZones):
    '''
    Write an array into a binary file
    '''
    mat = np.append(aantalZones, mat)
    matBin = array.array('f')
    matBin.fromlist(list(mat))
    matBin.tofile(open(filename, 'wb'))


def read_shape(shapePath, encoding='latin1', returnGeometry=False):
    '''
    Read the shapefile with zones (using pyshp --> import shapefile as shp)
    '''
    # Load the shape
    sf = shp.Reader(shapePath, encoding=encoding)
    records = sf.records()
    if returnGeometry:
        geometry = sf.__geo_interface__
        geometry = geometry['features']
        geometry = [geometry[i]['geometry'] for i in range(len(geometry))]
    fields = sf.fields
    sf.close()

    # Get information on the fields in the DBF
    columns = [x[0] for x in fields[1:]]
    colTypes = [x[1:] for x in fields[1:]]
    nRecords = len(records)

    # Check for headers that appear twice
    for col in range(len(columns)):
        name = columns[col]
        whereName = [i for i in range(len(columns)) if columns[i] == name]
        if len(whereName) > 1:
            for i in range(1, len(whereName)):
                columns[whereName[i]] = (
                    str(columns[whereName[i]]) + '_' + str(i))

    # Put all the data records into a NumPy array
    # (much faster than Pandas DataFrame)
    shape = np.zeros((nRecords, len(columns)), dtype=object)
    for i in range(nRecords):
        shape[i, :] = records[i][0:]

    # Then put this into a Pandas DataFrame with
    # the right headers and data types
    shape = pd.DataFrame(shape, columns=columns)
    for col in range(len(columns)):
        if colTypes[col][0] == 'C':
            shape[columns[col]] = shape[columns[col]].astype(str)
        else:
            shape.loc[pd.isna(shape[columns[col]]), columns[col]] = -99999
            if colTypes[col][-1] > 0:
                shape[columns[col]] = shape[columns[col]].astype(float)
            else:
                shape[columns[col]] = shape[columns[col]].astype(int)

    if returnGeometry:
        return (shape, geometry)
    else:
        return shape