我有逻辑回归模式,我明确地将阈值设置为0.5.
model.setThreshold(0.5)
我训练模型然后我想得到基本的统计数据 - 精确度,召回等.
这是我在评估模型时所做的事情:
val metrics = new BinaryClassificationMetrics(predictionAndLabels)
val precision = metrics.precisionByThreshold
precision.foreach { case (t, p) =>
println(s"Threshold is: $t, Precision is: $p")
}
我得到的结果只有0.0和1.0作为阈值,0.5完全被忽略.
以下是上述循环的输出:
阈值为:1.0,精度为:0.8571428571428571
阈值为:0.0,精度为:0.3005181347150259
当我调用metrics.thresholds()时,它也只返回两个值,0.0和1.0.
如何获得阈值为0.5的精度和召回值？
我需要这样做:
paste file1 file2 file3 > result
我的python脚本中有以下内容:
from subprocess import call
// other code here.
// Here is how I call the shell command
call ["paste", "file1", "file2", "file3", ">", "result"])
不幸的是我收到此错误:
paste: >: No such file or directory.
任何帮助都会很棒!
我有一个此数据类型的Spark RDD：RDD [（Int，Array [Int]）]）
该RDD的样本值为：
100，数组（1,2,3,4,5）
200，数组（1,2,50,20）
300，数组（30,2,400,1）
我想在此RDD的所有Array元素中获取所有唯一值，我不在乎键，只想获取所有唯一值。因此，以上示例的结果为（1,2,3,4,5,20,30,50,400）。
什么是这样做的有效方法。
val reslist = List(200.0,-100.00,50.80,-400.83, 800.003,-6.513214114672146E85, -1.2425461624057028E86, -4.7624471630469706E86, -3.6046499228286203E86, 0.0, -8.833653923554989E85, 0.000, -4.795843631190487E85, -5.34142100270833E86, -3.48087737474366E85, -2.811146396971388E86, -6.923235225460886E86, -6.513214114672146E85, 0.00000, -1.2425461624057028E86, -7.073704018243951E85, -9.633244016491059E86, -1.1418901590222212E86, -2.115257701350766E86, -1.1418901590222212E86, -3.48087737474366E85,-1.0676381955303372E86,500.56, 2.900556,400.56,-48956.00,4509.0005);
val weightlistzi = reslist.zipWithIndex
// List((200.0,0), (-100.0,1), (50.8,2), (-400.83,3), (800.003,4), (-6.513214114672146E85,5), (-1.2425461624057028E86,6), (-4.7624471630469706E86,7), (-3.6046499228286203E86,8), (0.0,9), (-8.833653923554989E85,10), (0.0,11), (-4.795843631190487E85,12), (-5.34142100270833E86,13), (-3.48087737474366E85,14), (-2.811146396971388E86,15), (-6.923235225460886E86,16), (-6.513214114672146E85,17), (0.0,18), (-1.2425461624057028E86,19), (-7.073704018243951E85,20), (-9.633244016491059E86,21), (-1.1418901590222212E86,22), (-2.115257701350766E86,23), (-1.1418901590222212E86,24), (-3.48087737474366E85,25), (-1.0676381955303372E86,26), (500.56,27), (2.900556,28), (400.56,29), (-48956.0,30), (4509.0005,31))
// I am sorting it here.
val resultlist = weightlistzi.sortWith { (x: (Double,Int), y: (Double,Int)) => x._1 …