What would be the best way to parse a text file that contains python dictionary?

4 次查看（过去 30 天）
Emmanuel
Emmanuel 2017-3-14
评论： Guillaume 2017-3-15
I have a text file that contains the following dictionary. I want to plot a bar graph of these values against given keys.
  1. How can I read a file that is in dictionary format?
  2. What is the best way to parse it in matlab? My Matlab version is R2015b
{"left_w0": -3.0177237015197758, "left_w1": -1.568878849017334, "left_w2": 1.2164467634033205, "right_s0": -0.5514660926147461, "right_s1": 0.2688301327697754, "right_w0": -1.9677138534118654, "right_w1": 1.4768400018493653, "head_pan": -0.04525243319091797, "right_w2": 2.27834496260376, "head_nod": 0.0, "torso_t0": -12.565987104803467, "left_e0": -0.996704015789795, "left_e1": 0.7044806760314942, "left_s0": -0.4613447214294434, "left_s1": -0.6074563913085937, "right_e0": 1.8457623809143067, "right_e1": 1.5876701136474611}
I want my output to just have the arrays of these numbers.

请先登录，再回答此问题。

回答（2 个）

Guillaume
Guillaume 2017-3-14
That string is JSON, which matlab can decode since R2016b (and R2017a fixes some of the bugs) with jsondecode. This will return a structure which can be easily converted into a containers.Map, the equivalent of a dictionary:
s = '{"left_w0": -3.0177237015197758, "left_w1": -1.568878849017334, "left_w2": 1.2164467634033205, "right_s0": -0.5514660926147461, "right_s1": 0.2688301327697754, "right_w0": -1.9677138534118654, "right_w1": 1.4768400018493653, "head_pan": -0.04525243319091797, "right_w2": 2.27834496260376, "head_nod": 0.0, "torso_t0": -12.565987104803467, "left_e0": -0.996704015789795, "left_e1": 0.7044806760314942, "left_s0": -0.4613447214294434, "left_s1": -0.6074563913085937, "right_e0": 1.8457623809143067, "right_e1": 1.5876701136474611}';
kvp = jsondecode(s)
mymap = containers.Map(fieldnames(kvp), struct2cell(kvp))
  2 个评论
Emmanuel
Emmanuel 2017-3-14
Thank you for your answer. My matlab version is R2015a. Also, My outpout should be an array of these numbers. Ijust edited my question..sorry about that.
Guillaume
Guillaume 2017-3-15
jsonlab on the FEX should allow you to decode JSON. It'll be safer than writing your own limited parser. There are several other submissions there as well (e.g. this one)
You can then easily convert whatever these output into an array.

请先登录，再进行评论。

Walter Roberson
Walter Roberson 2017-3-14
s = '{"left_w0": -3.0177237015197758, "left_w1": -1.568878849017334, "left_w2": 1.2164467634033205, "right_s0": -0.5514660926147461, "right_s1": 0.2688301327697754, "right_w0": -1.9677138534118654, "right_w1": 1.4768400018493653, "head_pan": -0.04525243319091797, "right_w2": 2.27834496260376, "head_nod": 0.0, "torso_t0": -12.565987104803467, "left_e0": -0.996704015789795, "left_e1": 0.7044806760314942, "left_s0": -0.4613447214294434, "left_s1": -0.6074563913085937, "right_e0": 1.8457623809143067, "right_e1": 1.5876701136474611}';
parts = regexp(s(2:end-1), '[,:]\s*', 'split');
str2double(parts(2:2:end))
  3 个评论
显示 1更早的评论
Walter Roberson
Walter Roberson 2017-3-15
You did not happen to post a sample file.
I speculate that the JSON is UTF encoded; I think I have seen that on some JSON.
I refer you to https://www.mathworks.com/matlabcentral/answers/270857-how-to-read-16-bit-text-with-matlab and in particular to the link there to the source code I posted to detect UTF encoding, which is a lot of the struggle in reading an unknown file.
Emmanuel
Emmanuel 2017-3-15
Yeah sure! Thank you very much. What I did was, I parsed it in python and converted it into an array of numbers and stored it in text file. I will definitely refer to your link as well. Thank you!

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Text Files 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by