DataFrame

Matlab impelementation of DataFrame/Pandas concept. This project wraps and makes use of as much as possible of the Matlab table, but with the intent of providing a class implementation that could be specialized further.

</style>

Create a DataFrame from scratch

The DataFrame is created in the same exact way as a table. The constructor just passes the arguements to the table constructor, then stores the table as as private property. We can still retrieve table properties

df = DataFrame(0, 0, 0, 0, ...'VariableNames', {'test1', 'test2','test3','test4',});
df.Properties

ans =
 Description: ''
VariableDescriptions: {}
VariableUnits: {}
DimensionNames: {'Row' 'Variable'}
UserData: []
RowNames: {}
VariableNames: {'test1' 'test2' 'test3' 'test4'}

View the DataFrame the Same as a Table

DataFrame overrides the display method to show the table instead of the DataFrame object

df

df =
test1 test2 test3 test4
_____ _____ _____ _____
0 0 0 0 

Access Table Columns Through DataFrame

The column names are pass throughs into the actual table structure. We can access the column as normal

df.test1
df.test3

ans =
 0
ans =
 0

Access Methods Attached to the DataFrame Object

You no longer need to know which possible functions you can use on a Matlab table. They are available to use either way.

df.height()
height(df)

ans =
 1
ans =
 1

Extend DataFrame However We'd Like

Matlab blocks you from extending the table data structure, but we can get around that with this approach

df = DataFrame.fromCSV(which('ugly_data.csv'));
df.head();

 Name Zeros Lat Lng Normal Negative
_______________ _____ ________ ________ ___________ ________
'Connor Hayden' 0 -2.58201 71.06714 0.006049531 3 GUID Date Timestamp ______________________________________ _________________ __________
'7B79A197-E85F-2BFF-F37E-727DEDAC9803' 'April 27th 2015' 1394414933
AlphaNumeric _____________
'FSU09MJG3ZB'

Or, if You Still Want to See the Object

Column data will show empty since we are just dynamically passing through them on each subscript call. Properties shows up with contents, since it uses a getter method

df.details();

 DataFrame with properties:
 Properties: [1x1 struct]
Normal: []
Lng: []
Negative: []
Zeros: []
Name: []
AlphaNumeric: []
Date: []
GUID: []
Lat: []
Timestamp: []
Name Zeros Lat Lng Normal Negative
_______________ _____ ________ ________ ___________ ________
'Connor Hayden' 0 -2.58201 71.06714 0.006049531 3 GUID Date Timestamp ______________________________________ _________________ __________
'7B79A197-E85F-2BFF-F37E-727DEDAC9803' 'April 27th 2015' 1394414933
AlphaNumeric _____________
'FSU09MJG3ZB'
Variables:
Name: 20400x1 cell string
Zeros: 20400x1 double
Values:
min 0 median 0 max 0 Lat: 20400x1 double
Values:
min -89.42406
median -1.02059
max 88.89347
Lng: 20400x1 double
Values:
min -177.18611
median 0.934815
max 176.4439
Normal: 20400x1 double
Values:
min -0.543252895
median -0.017611593
max 0.573277103
Negative: 20400x1 double
Values:
min -5 median 0.5 max 5 GUID: 20400x1 cell string
Date: 20400x1 cell string
Timestamp: 20400x1 double
Values:
min 1378967930
median 1403645750.5
max 1439505058
AlphaNumeric: 20400x1 cell string

We Can Try the Same Thing on a Standard Table

tbl = readtable(which('ugly_data.csv'));
try
tbl.head();
head(tbl);
catch err
disp(err.message)
end

Unrecognized variable name 'head'.

Dynamically add columns to the DataFrame

df.Lat2 = df.Lat;
df.Lat3 = df.Lat;
df.columns()

ans =
Columns 1 through 7
'Name' 'Zeros' 'Lat' 'Lng' 'Normal' 'Negative' 'GUID'
Columns 8 through 12
'Date' 'Timestamp' 'AlphaNumeric' 'Lat2' 'Lat3'

Easily remove columns from the DataFrame

df.remove_cols({'Lat2', 'Lat3'});
df.columns()

ans =
Columns 1 through 7
'Name' 'Zeros' 'Lat' 'Lng' 'Normal' 'Negative' 'GUID'
Columns 8 through 10
'Date' 'Timestamp' 'AlphaNumeric'

Initial Start at Providing Complete Wrapper for Table

methods('DataFrame')

Methods for class DataFrame:
DataFrame disp is_column subsref toStruct

addprop getTable numel summary width

columns head remove_cols toArray writetable

details height subsasgn toCell
Static methods:
fromArray fromCell intersect rowfun

fromCSV fromStruct ismember varfun
Call "methods('handle')" for methods of DataFrame inherited from handle.