CORNELLSASLAB

SAS output to EXCEL converter for Cornell/MIT Language and acquisition lab

Instructions:

This python code can be used to convert SAS outputs from the lang and ac lab into data tables for regression and factorial summaries. SAS formst varies by run, therefore you may have to do a little editing before the code works.

For example, a working data chunk for this program would look something like this:

                  Differences of CHILDAGE*TASK*TR Least Squares Means

                                                     Standard
CHILDAGE  TASK  TR  _CHILDAGE  _TASK  _TR  Estimate     Error     DF  t Value  Pr > |t|

1         2     1   8          1      1     -0.4167    0.1251   1544    -3.33    0.0009
1         2     1   8          1      2     -0.2917    0.1251   1544    -2.33    0.0199
1         2     2   7          2      3     -0.4375    0.1363   1544    -3.21    0.0014
1         2     2   8          1      1     -0.4583    0.1251   1544    -3.66    0.0003

The SAS System 15:01 Sunday, August 4, 2019 139

                                 The GLIMMIX Procedure

                  Differences of CHILDAGE*TASK*TR Least Squares Means

                                                     Standard
CHILDAGE  TASK  TR  _CHILDAGE  _TASK  _TR  Estimate     Error     DF  t Value  Pr > |t|

1         2     2   8          1      2     -0.3333    0.1251   1544    -2.66    0.0078
1         2     2   8          1      3     -0.5625    0.1251   1544    -4.50    <.0001
1         2     2   8          2      1     -0.5833    0.1363   1544    -4.28    <.0001
1         2     3   7          2      3     -0.4167    0.1363   1544    -3.06    0.0023
1         2     3   8          1      1     -0.4375    0.1251   1544    -3.50    0.0005
1         2     3   8          1      2     -0.3125    0.1251   1544    -2.50    0.0126
1         2     3   8          1      3     -0.5417    0.1251   1544    -4.33    <.0001

The SAS System 15:01 Sunday, August 4, 2019 140

                                 The GLIMMIX Procedure

                  Differences of CHILDAGE*TASK*TR Least Squares Means

                                                     Standard
CHILDAGE  TASK  TR  _CHILDAGE  _TASK  _TR  Estimate     Error     DF  t Value  Pr > |t|

1         2     3   8          2      1     -0.5625    0.1363   1544    -4.13    <.0001
1         2     3   8          2      2     -0.5208    0.1363   1544    -3.82    0.0001
1         2     3   8          2      3     -0.4583    0.1363   1544    -3.36    0.0008
2         1     1   2          1      2    6.66E-16   0.07654   1544     0.00    1.0000
2         1     1   2          1      3    -0.08333   0.07654   1544    -1.09    0.2764
2         1     1   8          2      1     -0.4583    0.1251   1544    -3.66    0.0003
2         1     1   8          2      2     -0.4167    0.1251   1544    -3.33    0.0009
2         1     1   8          2      3     -0.3542    0.1251   1544    -2.83    0.0047

The SAS System 15:01 Sunday, August 4, 2019 141

As you can see, every chunk starts with the same title, and ends with a date and page number in the bottom right corner. All of the chunks have the same format and same number of columns. If the format of the pasted chunks is wrong, the code likely won't work.

NOTE: when pasting the file path to your excel document, make sure excel is not running on your computer, or else there will be an error.

SAS output to EXCEL converter for Cornell/MIT Language and acquisition lab

Related tags

Overview

CORNELLSASLAB

Owner

A novel pipeline framework for multi-hop complex KGQA task. About the paper title: Improving Multi-hop Embedded Knowledge Graph Question Answering by Introducing Relational Chain Reasoning

Code for project: "Learning to Minimize Remainder in Supervised Learning".

Code for the paper "Reinforced Active Learning for Image Segmentation"

A minimalist implementation of score-based diffusion model

Official Implementation of CVPR 2022 paper: "Mimicking the Oracle: An Initial Phase Decorrelation Approach for Class Incremental Learning"

Enhancing Knowledge Tracing via Adversarial Training

ViDT: An Efficient and Effective Fully Transformer-based Object Detector

Reinforcement Learning via Supervised Learning

This is an implementation of PIFuhd based on Pytorch

Code to reproduce the experiments in the paper "Transformer Based Multi-Source Domain Adaptation" (EMNLP 2020)

DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting

Generate high quality pictures. GAN. Generative Adversarial Networks

Simple Tensorflow implementation of "Adaptive Convolutions for Structure-Aware Style Transfer" (CVPR 2021)

Supervised Classification from Text (P)

Implementation and replication of ProGen, Language Modeling for Protein Generation, in Jax

code for our BMVC 2021 paper "HCV: Hierarchy-Consistency Verification for Incremental Implicitly-Refined Classification"

maximal update parametrization (µP)

Avatarify Python - Avatars for Zoom, Skype and other video-conferencing apps.

This is a JAX implementation of Neural Radiance Fields for learning purposes.

Really awesome semantic segmentation