synw / df Goto Github PK

Currently all columns have nullable data types. If a user has data that explicitly should not be null, this makes their code more verbose (they have to add ! everywhere) and doesn't enforce non-null input.

To support non-nullable data types, we should add a non-nullable column and/or non-nullable DataFrame.

I can create a tmp file to prove that the asset is working, but the .fromCsv function appears to break. What am I doing wrong here?

Future<void> loadUSZIPCodesCSV() async {
String path = 'assets/files/uszips.csv';
final file = File(path);
var existsSync = file.existsSync();
print('File path is actually: ${file.path} and ${existsSync}');
final df = await DataFrame.fromCsv(file.path,
dateFormat: "MMM dd yyyy", verbose: true);
df.show();
print(df.columns);
}

My log is:

I/flutter (24093): File path is actually: assets/files/uszips.csv and false
E/flutter (24093): [ERROR:flutter/lib/ui/ui_dart_state.cc(157)] Unhandled Exception: Instance of 'FileNotFoundException'
E/flutter (24093): #0      DataFrame.fromCsv (package:df/src/df.dart:124:7)

The following code doesn't work on android. While the system prints 'exists' which means I'm using the correct path, the df package throws
[ERROR:flutter/lib/ui/ui_dart_state.cc(209)] Unhandled Exception: Instance of 'FileNotFoundException'

The same code works on windows. Obviously has permission to read files as it prints 'exists' as well.

if (!File('downloaded/filters/merged/test.csv').existsSync()) {
  print('exists');
}
final df = await DataFrame.fromCsv('downloaded/filters/merged/test.csv');

Google sheets and many other spreadsheet editing tools escape commas contained in cell values by wrapping the record in double quotes.

DF currently throws a Range Error if a row contains a record with a double quote escaped comma. DF should respect double quote escaping.

I need the feature to convert html table tag to dataframe like pandas one

Is there a way of converting a df to a csv file or some other way of saving it.

This library does not have any native dependencies and can be used across all platforms. Adding the web and desktop tags may help with exposure of this great library. Thanks for building this!

Expected behaviour:

df.drop('columnName')
df.dropna() //drops columns containing empty values

Add how to perform Python like action:

df = pd.DataFrame(csvPath)
filteredDf = df[df['price']=15]

As of now seems it is not possible to perform such action (Select {columns} Where {condition})