public override DataFrameColumn Add(DataFrameColumn column, bool inPlace = false)
{
if (Length != column.Length)
{
throw new ArgumentException(Strings.MismatchedColumnLengths, nameof(column));
}
StringDataFrameColumn ret = inPlace ? this : Clone();
for (long i = 0; i < Length; i++)
{
ret[i] += column[i].ToString();
}
return(ret);
}
public StringDataFrameColumn Add(string value, bool inPlace = false)
{
StringDataFrameColumn ret = inPlace ? this : Clone();
for (int i = 0; i < ret._stringBuffers.Count; i++)
{
IList <string> buffer = ret._stringBuffers[i];
int bufferLen = buffer.Count;
for (int j = 0; j < bufferLen; j++)
{
buffer[j] += value;
}
}
return(ret);
}
public static StringDataFrameColumn Add(string value, StringDataFrameColumn right)
{
StringDataFrameColumn ret = right.Clone();
for (int i = 0; i < ret._stringBuffers.Count; i++)
{
IList <string> buffer = ret._stringBuffers[i];
int bufferLen = buffer.Count;
for (int j = 0; j < bufferLen; j++)
{
buffer[j] = value + buffer[j];
}
}
return(ret);
}
private static DataFrameColumn CreateColumn(Type kind, string[] columnNames, int columnIndex)
{
PrimitiveDataFrameColumn <T> CreatePrimitiveDataFrameColumn <T>()
where T : unmanaged
{
return(new PrimitiveDataFrameColumn <T>(columnNames == null ? "Column" + columnIndex.ToString() : columnNames[columnIndex]));
}
DataFrameColumn ret;
if (kind == typeof(bool))
{
ret = CreatePrimitiveDataFrameColumn <bool>();
}
else if (kind == typeof(int))
{
ret = CreatePrimitiveDataFrameColumn <int>();
}
else if (kind == typeof(float))
{
ret = CreatePrimitiveDataFrameColumn <float>();
}
else if (kind == typeof(string))
{
ret = new StringDataFrameColumn(columnNames == null ? "Column" + columnIndex.ToString() : columnNames[columnIndex], 0);
}
else if (kind == typeof(long))
{
ret = CreatePrimitiveDataFrameColumn <long>();
}
else if (kind == typeof(decimal))
{
ret = CreatePrimitiveDataFrameColumn <decimal>();
}
else if (kind == typeof(byte))
{
ret = CreatePrimitiveDataFrameColumn <byte>();
}
else if (kind == typeof(char))
{
ret = CreatePrimitiveDataFrameColumn <char>();
}
else if (kind == typeof(double))
{
ret = CreatePrimitiveDataFrameColumn <double>();
}
else if (kind == typeof(sbyte))
{
ret = CreatePrimitiveDataFrameColumn <sbyte>();
}
else if (kind == typeof(short))
{
ret = CreatePrimitiveDataFrameColumn <short>();
}
else if (kind == typeof(uint))
{
ret = CreatePrimitiveDataFrameColumn <uint>();
}
else if (kind == typeof(ulong))
{
ret = CreatePrimitiveDataFrameColumn <ulong>();
}
else if (kind == typeof(ushort))
{
ret = CreatePrimitiveDataFrameColumn <ushort>();
}
else
{
throw new NotSupportedException(nameof(kind));
}
return(ret);
}
private static DataFrameColumn CreateColumn(Type kind, string[] columnNames, int columnIndex)
{
DataFrameColumn ret;
if (kind == typeof(bool))
{
ret = new BooleanDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(int))
{
ret = new Int32DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(float))
{
ret = new SingleDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(string))
{
ret = new StringDataFrameColumn(GetColumnName(columnNames, columnIndex), 0);
}
else if (kind == typeof(long))
{
ret = new Int64DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(decimal))
{
ret = new DecimalDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(byte))
{
ret = new ByteDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(char))
{
ret = new CharDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(double))
{
ret = new DoubleDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(sbyte))
{
ret = new SByteDataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(short))
{
ret = new Int16DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(uint))
{
ret = new UInt32DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(ulong))
{
ret = new UInt64DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else if (kind == typeof(ushort))
{
ret = new UInt16DataFrameColumn(GetColumnName(columnNames, columnIndex));
}
else
{
throw new NotSupportedException(nameof(kind));
}
return(ret);
}
/// <summary>
/// Reads a seekable stream of CSV data into a DataFrame.
/// Follows pandas API.
/// </summary>
/// <param name="csvStream">stream of CSV data to be read in</param>
/// <param name="separator">column separator</param>
/// <param name="header">has a header or not</param>
/// <param name="columnNames">column names (can be empty)</param>
/// <param name="dataTypes">column types (can be empty)</param>
/// <param name="numberOfRowsToRead">number of rows to read not including the header(if present)</param>
/// <param name="guessRows">number of rows used to guess types</param>
/// <param name="addIndexColumn">add one column with the row index</param>
/// <returns>DataFrame</returns>
public static DataFrame LoadCsv(Stream csvStream,
char separator = ',', bool header = true,
string[] columnNames = null, Type[] dataTypes = null,
long numberOfRowsToRead = -1, int guessRows = 10, bool addIndexColumn = false)
{
if (!csvStream.CanSeek)
{
throw new ArgumentException(Strings.NonSeekableStream, nameof(csvStream));
}
var linesForGuessType = new List <string[]>();
long rowline = 0;
int numberOfColumns = 0;
if (header == true && numberOfRowsToRead != -1)
{
numberOfRowsToRead++;
}
List <DataFrameColumn> columns;
long streamStart = csvStream.Position;
// First pass: schema and number of rows.
using (var streamReader = new StreamReader(csvStream, encoding: null, detectEncodingFromByteOrderMarks: true, bufferSize: -1, leaveOpen: true))
{
string line = streamReader.ReadLine();
while (line != null)
{
if ((numberOfRowsToRead == -1) || rowline < numberOfRowsToRead)
{
if (linesForGuessType.Count < guessRows)
{
var spl = line.Split(separator);
if (header && rowline == 0)
{
if (columnNames == null)
{
columnNames = spl;
}
}
else
{
linesForGuessType.Add(spl);
numberOfColumns = Math.Max(numberOfColumns, spl.Length);
}
}
}
++rowline;
if (rowline == numberOfRowsToRead)
{
break;
}
line = streamReader.ReadLine();
}
if (linesForGuessType.Count == 0)
{
throw new FormatException(Strings.EmptyFile);
}
columns = new List <DataFrameColumn>(numberOfColumns);
// Guesses types and adds columns.
for (int i = 0; i < numberOfColumns; ++i)
{
Type kind = GuessKind(i, linesForGuessType);
if (kind == typeof(bool))
{
DataFrameColumn boolColumn = new PrimitiveDataFrameColumn <bool>(columnNames == null ? "Column" + i.ToString() : columnNames[i], header == true ? rowline - 1 : rowline);
columns.Add(boolColumn);
}
else if (kind == typeof(float))
{
DataFrameColumn floatColumn = new PrimitiveDataFrameColumn <float>(columnNames == null ? "Column" + i.ToString() : columnNames[i], header == true ? rowline - 1 : rowline);
columns.Add(floatColumn);
}
else if (kind == typeof(string))
{
DataFrameColumn stringColumn = new StringDataFrameColumn(columnNames == null ? "Column" + i.ToString() : columnNames[i], header == true ? rowline - 1 : rowline);
columns.Add(stringColumn);
}
else
{
throw new NotSupportedException(nameof(kind));
}
}
line = null;
streamReader.DiscardBufferedData();
streamReader.BaseStream.Seek(streamStart, SeekOrigin.Begin);
// Fills values.
line = streamReader.ReadLine();
rowline = 0;
while (line != null && (numberOfRowsToRead == -1 || rowline < numberOfRowsToRead))
{
var spl = line.Split(separator);
if (header && rowline == 0)
{
// Skips.
}
else
{
AppendRow(columns, header == true ? rowline - 1 : rowline, spl);
}
++rowline;
line = streamReader.ReadLine();
}
if (addIndexColumn)
{
PrimitiveDataFrameColumn <int> indexColumn = new PrimitiveDataFrameColumn <int>("IndexColumn", columns[0].Length);
for (int i = 0; i < columns[0].Length; i++)
{
indexColumn[i] = i;
}
columns.Insert(0, indexColumn);
}
}
return(new DataFrame(columns));
}
