Design the Mechanism of Writing the Output of Random Forest Algorithm into an Output File

[azmfaridee]

So far we have been printing the output in stdout, we need to be able to write the output of the algorithm to an output file just like other mothur commands.

@kdiverson @mothur-westcott Can you give me an idea about this?

Out current debugging output that we print to stdout is somewhat like this

numCorrect 150
forrestErrorRate: 0.197860962567
globalVariableRanks: [[9, 5.7], [2, 0.68], [44, 0.19], [158, 0.17], [14, 0.15], [27, 0.15], [31, 0.15], [144, 0.15], [33, 0.14], [16, 0.13], [86, 0.13], [141, 0.13], [182, 0.13], [264, 0.12], [41, 0.11], [302, 0.11], [36, 0.1], [58, 0.09], [161, 0.09], [22, 0.08], [89, 0.08], [145, 0.08], [360, 0.08], [11, 0.07], [37, 0.06], [40, 0.06], [56, 0.06], [60, 0.06], [71, 0.06], [105, 0.06], [133, 0.06], [157, 0.06], [243, 0.06], [285, 0.06], [1, 0.05], [21, 0.05], [23, 0.05], [24, 0.05], [47, 0.05], [68, 0.05], [82, 0.05], [98, 0.05], [117, 0.05], [195, 0.05], [286, 0.05], [568, 0.05], [15, 0.04], [30, 0.04], [51, 0.04], [72, 0.04], [91, 0.04], [114, 0.04], [127, 0.04], [129, 0.04], [184, 0.04], [230, 0.04], [335, 0.04], [7, 0.03], [26, 0.03], [64, 0.03], [69, 0.03], [70, 0.03], [84, 0.03], [151, 0.03], [154, 0.03], [175, 0.03], [176, 0.03], [202, 0.03], [258, 0.03], [293, 0.03], [296, 0.03], [427, 0.03], [3, 0.02], [35, 0.02], [74, 0.02], [81, 0.02], [87, 0.02], [92, 0.02], [93, 0.02], [112, 0.02], [116, 0.02], [140, 0.02], [164, 0.02], [167, 0.02], [180, 0.02], [207, 0.02], [215, 0.02], [219, 0.02], [231, 0.02], [266, 0.02], [267, 0.02], [338, 0.02], [346, 0.02], [399, 0.02], [400, 0.02], [407, 0.02], [562, 0.02], [643, 0.02], [12, 0.01], [17, 0.01], [29, 0.01], [32, 0.01], [54, 0.01], [61, 0.01], [65, 0.01], [67, 0.01], [73, 0.01], [88, 0.01], [100, 0.01], [102, 0.01], [113, 0.01], [115, 0.01], [119, 0.01], [142, 0.01], [143, 0.01], [159, 0.01], [183, 0.01], [218, 0.01], [224, 0.01], [252, 0.01], [256, 0.01], [265, 0.01], [273, 0.01], [276, 0.01], [277, 0.01], [280, 0.01], [287, 0.01], [294, 0.01], [308, 0.01], [334, 0.01], [339, 0.01], [371, 0.01], [382, 0.01], [406, 0.01], [409, 0.01], [421, 0.01], [444, 0.01], [468, 0.01], [532, 0.01], [673, 0.01], [737, 0.01]]

We need to design the format of the output file for this case. I'm thinking of a summary message followed by two column data. e.g.

Error Rate:    19.79%
--------------------------------------
Summary of Variable Ranks
--------------------------------------
OTU                 RANK
--------------------------------------
OTU10            5.7
OTU3              0.66
...                      ....
...                      ....

@kdiverson @mothur-westcott What do you think? Anything else we need to add in the output file. Also, I think each of the runs with different parameters would be different, therefore, re-using the output result from a previous runs wold be unnecessary (e.g. the forest created with log based criteria and the forest created with square-root based criteria would be different, and there would be no way to use one runs data for another run)

[mothur-westcott]

If you are going to output the OTU labels be sure to use the labels from m->currentBinLabels. They are set when the sharedRabundVector class reads the shared file. Other than that, I will defer to Kathryn's expertise as to what kind of output would be most helpful to users.

[kdiverson]

That output looks good.