Data preparation for susceptibility mapping-Part II

import pandas as pd

aspect=pd.read_csv('/work/small_aspect.csv') aspect.head()

Rowid_float64

ASPECT3_TIFint64

nan

256

nan

aspect=aspect.drop(['ASPECT3_TIF','Rowid_'], axis = 1) #drop extra column

aspect.head()

Xfloat64

Yfloat64

431720.9935075852

3987844.4410662935

431749.0169511396

3987844.4410662935

431749.0169511396

3987816.417622739

431805.0638382485

3987844.4410662935

431833.0872818029

3987844.4410662935

aspect.columns=['X','Y','aspect'] #rename columns

aspect['ID'] = aspect.index # add new index column

aspect.head()

Xfloat64

Yfloat64

431720.9935075852

3987844.4410662935

431749.0169511396

3987844.4410662935

431749.0169511396

3987816.417622739

431805.0638382485

3987844.4410662935

431833.0872818029

3987844.4410662935

aspect = aspect[['ID','X','Y','aspect']] # rearrange columns

aspect.to_csv ('/work/aspect.csv', index=None) # save back to csv

dem=pd.read_csv('/work/small_dem.csv') dem.head()

Rowid_float64

DEMint64

nan

3445

nan

3440

nan

3442

nan

3398

nan

3388

dem=dem.drop(['Rowid_','DEM'], axis = 1)

dem.columns=['X','Y','DEM']

dem['ID'] = dem.index

dem.head()

Xfloat64

Yfloat64

431720.9935075852

3987844.4410662935

431749.0169511396

3987844.4410662935

431749.0169511396

3987816.417622739

431805.0638382485

3987844.4410662935

431833.0872818029

3987844.4410662935

dem = dem[['ID','X','Y','DEM']]

dem.head()

IDint64

Xfloat64

431720.9935075852

431749.0169511396

431805.0638382485

431833.0872818029

dem.to_csv ('/work/dem.csv', index=None)

lulc=pd.read_csv('/work/small_lulc.csv') lulc.head()

Rowid_float64

LULC1int64

nan

lulc=lulc.drop(['Rowid_','LULC1'], axis = 1)

lulc['ID'] = lulc.index

lulc.head()

Xfloat64

Yfloat64

431720.9935075852

3987844.4410662935

431749.0169511396

3987844.4410662935

431749.0169511396

3987816.417622739

431805.0638382485

3987844.4410662935

431833.0872818029

3987844.4410662935

lulc = lulc[['ID','X','Y','LULC1_1']]

lulc.columns=['ID','X','Y','LULC']

lulc.head()

IDint64

Xfloat64

431720.9935075852

431749.0169511396

431805.0638382485

431833.0872818029

lulc.to_csv ('/work/lulc.csv', index=None)

#open all factors again from their saved modified csv files aspect=pd.read_csv('/work/aspect.csv')

dem=pd.read_csv('/work/dem.csv')

lulc=pd.read_csv('/work/lulc.csv')

from functools import reduce

#combine all factors data frames based on ID column df_prediction = reduce(lambda x,y: pd.merge(x,y, on='ID', how='outer'), [aspect, dem, lulc]) df_prediction.head()

IDint64

X_xfloat64

431720.9935075852

431749.0169511396

431805.0638382485

431833.0872818029

#we got multiple X, Y columns so drop extra one's df_prediction=df_prediction.drop(['X_x','Y_y','X_y','Y_x'], axis = 1)

df_prediction

IDint64

0 - 9

aspectfloat64

38.4801979 - 256.6075134

256.6075134

76.7594833

76.6075058

52.4314079

38.4801979

49.9117889

58.5431747

61.9797096

55.0302124

40.6777229

#rearranging columns according to training data frame (created in part I) #here I gave example of just 3 factors so it is different from training data frame which include all factors df_prediction=df_prediction[['ID','X','Y','aspect','DEM','LULC']]

df_prediction.head()

IDint64

Xfloat64

431720.9935075852

431749.0169511396

431805.0638382485

431833.0872818029

df_prediction.to_csv ('/work/df_prediction.csv', index=None) # save dataframe

df_prediction=pd.read_csv('df_prediction.csv') #read csv file